Fixed issue for punctuation (#1344)

https://github.com/jambonz/jambonz-feature-server/issues/1343

app.js

@@ -29,6 +29,12 @@ const {LifeCycleEvents, FS_UUID_SET_NAME, SystemState, FEATURE_SERVER} = require
 const installSrfLocals = require('./lib/utils/install-srf-locals');
 const createHttpListener = require('./lib/utils/http-listener');
 const healthCheck = require('@jambonz/http-health-check');
+const ProcessMonitor = require('./lib/utils/process-monitor');
+const monitor = new ProcessMonitor(logger);
+
+// Log startup
+monitor.logStartup();
+monitor.setupSignalHandlers();
 
 logger.on('level-change', (lvl, _val, prevLvl, _prevVal, instance) => {
   if (logger !== instance) {

lib/config.js

@@ -130,7 +130,7 @@ const OPTIONS_PING_INTERVAL = parseInt(process.env.OPTIONS_PING_INTERVAL, 10) ||
 const JAMBONZ_RECORD_WS_BASE_URL = process.env.JAMBONZ_RECORD_WS_BASE_URL || process.env.JAMBONES_RECORD_WS_BASE_URL;
 const JAMBONZ_RECORD_WS_USERNAME = process.env.JAMBONZ_RECORD_WS_USERNAME || process.env.JAMBONES_RECORD_WS_USERNAME;
 const JAMBONZ_RECORD_WS_PASSWORD = process.env.JAMBONZ_RECORD_WS_PASSWORD || process.env.JAMBONES_RECORD_WS_PASSWORD;
-const JAMBONZ_DISABLE_DIAL_PAI_HEADER = process.env.JAMBONZ_DISABLE_DIAL_PAI_HEADER || false;
+const JAMBONZ_DIAL_PAI_HEADER = process.env.JAMBONZ_DIAL_PAI_HEADER || false;
 const JAMBONES_DISABLE_DIRECT_P2P_CALL = process.env.JAMBONES_DISABLE_DIRECT_P2P_CALL || false;
 
 const JAMBONES_EAGERLY_PRE_CACHE_AUDIO = parseInt(process.env.JAMBONES_EAGERLY_PRE_CACHE_AUDIO, 10) || 0;
@@ -225,11 +225,11 @@ module.exports = {
   JAMBONZ_RECORD_WS_BASE_URL,
   JAMBONZ_RECORD_WS_USERNAME,
   JAMBONZ_RECORD_WS_PASSWORD,
-  JAMBONZ_DISABLE_DIAL_PAI_HEADER,
+  JAMBONZ_DIAL_PAI_HEADER,
   JAMBONES_DISABLE_DIRECT_P2P_CALL,
   JAMBONES_USE_FREESWITCH_TIMER_FD,
   JAMBONES_DIAL_SBC_FOR_REGISTERED_USER,
   JAMBONES_MEDIA_TIMEOUT_MS,
   JAMBONES_MEDIA_HOLD_TIMEOUT_MS,
-  JAMBONES_TRANSCRIBE_EP_DESTROY_DELAY_MS
+  JAMBONES_TRANSCRIBE_EP_DESTROY_DELAY_MS,
 };

lib/session/call-session.js

@@ -10,8 +10,9 @@ const {
   RecordState,
   AllowedSipRecVerbs,
   AllowedConfirmSessionVerbs,
-  TtsStreamingEvents
-} = require('../utils/constants');
+  TtsStreamingEvents,
+  ListenStatus
+} = require('../utils/constants.json');
 const moment = require('moment');
 const assert = require('assert');
 const sessionTracker = require('./session-tracker');
@@ -37,6 +38,7 @@ const BADPRECONDITIONS = 'preconditions not met';
 const CALLER_CANCELLED_ERR_MSG = 'Response not sent due to unknown transaction';
 const { NonFatalTaskError} = require('../utils/error');
 const { createMediaEndpoint } = require('../utils/media-endpoint');
+const SttLatencyCalculator = require('../utils/stt-latency-calculator');
 const sqlRetrieveQueueEventHook = `SELECT * FROM webhooks 
 WHERE webhook_sid = 
 (
@@ -146,6 +148,30 @@ class CallSession extends Emitter {
     this.conversationTurns = [];
     this.on('userSaid', this._onUserSaid.bind(this));
     this.on('botSaid', this._onBotSaid.bind(this));
+    /**
+     * Support STT latency
+     */
+    this.sttLatencyCalculator = new SttLatencyCalculator({
+      logger,
+      cs: this
+    });
+    this.on('transcribe-start', () => {
+      this.sttLatencyCalculator.resetTime();
+    });
+    this.on('on-transcription', () => {
+      this.sttLatencyCalculator.onTranscriptionReceived();
+    });
+    this.on('transcribe-stop', () => {
+      this.sttLatencyCalculator.onTranscribeStop();
+    });
+  }
+
+  get notifySttLatencyEnabled() {
+    return this._notifySttLatencyEnabled || false;
+  }
+
+  set notifySttLatencyEnabled(enabled) {
+    this._notifySttLatencyEnabled = enabled;
   }
 
   /**
@@ -684,7 +710,7 @@ class CallSession extends Emitter {
   }
 
 
-  hasGlobalSttPunctuation() {
+  get hasGlobalSttPunctuation() {
     return this._globalSttPunctuation !== undefined;
   }
 
@@ -716,10 +742,25 @@ class CallSession extends Emitter {
     return this._fillerNoise;
   }
 
+  async pauseOrResumeBackgroundListenIfRequired(action, silence = false) {
+    if ((action == 'pauseCallRecording' || action == 'resumeCallRecording') &&
+    this.backgroundTaskManager.isTaskRunning('record')) {
+      this.logger.debug({action, silence}, 'CallSession:pauseOrResumeBackgroundListenIfRequired');
+      const backgroundListenTask = this.backgroundTaskManager.getTask('record');
+      const status = action === 'pauseCallRecording' ? ListenStatus.Pause : ListenStatus.Resume;
+      backgroundListenTask.updateListen(
+        status,
+        silence
+      );
+    }
+  }
+
   async notifyRecordOptions(opts) {
-    const {action} = opts;
+    const {action, silence} = opts;
     this.logger.debug({opts}, 'CallSession:notifyRecordOptions');
 
+    this.pauseOrResumeBackgroundListenIfRequired(action, silence);
+
     /* if we have not answered yet, just save the details for later */
     if (!this.dlg) {
       if (action === 'startCallRecording') {
@@ -976,8 +1017,6 @@ class CallSession extends Emitter {
         (type === 'tts' && credential.use_for_tts) ||
         (type === 'stt' && credential.use_for_stt)
       )) {
-        this.logger.debug(
-          `${type}: ${credential.vendor} ${credential.label ? `, label: ${credential.label}` : ''}  `);
         if ('google' === vendor) {
           if (type === 'tts' && !credential.tts_tested_ok ||
             type === 'stt' && !credential.stt_tested_ok) {
@@ -1105,6 +1144,13 @@ class CallSession extends Emitter {
             options: credential.options
           };
         }
+        else if ('resemble' === vendor) {
+          return {
+            api_key: credential.api_key,
+            resemble_tts_use_tls: credential.resemble_tts_use_tls,
+            resemble_tts_uri: credential.resemble_tts_uri,
+          };
+        }
         else if ('inworld' === vendor) {
           return {
             api_key: credential.api_key,
@@ -1119,6 +1165,12 @@ class CallSession extends Emitter {
             service_version: credential.service_version
           };
         }
+        else if ('deepgramriver' === vendor) {
+          return {
+            speech_credential_sid: credential.speech_credential_sid,
+            api_key: credential.api_key,
+          };
+        }
         else if ('voxist' === vendor) {
           return {
             speech_credential_sid: credential.speech_credential_sid,
@@ -1661,8 +1713,8 @@ Duration=${duration} `
   async _lccWhisper(opts, callSid) {
     const {whisper} = opts;
     let tasks;
-    const b3 = this.b3;
-    const httpHeaders = b3 && {b3};
+    //const b3 = this.b3;
+    //const httpHeaders = b3 && {b3};
 
     // this whole thing requires us to be in a Dial verb
     const task = this.currentTask;
@@ -1671,12 +1723,15 @@ Duration=${duration} `
     }
 
     // allow user to provide a url object, a url string, an array of tasks, or a single task
-    if (typeof whisper === 'string' || (typeof whisper === 'object' && whisper.url)) {
-      // retrieve a url
-      const json = await this.requestor(opts.call_hook, this.callInfo.toJSON(), httpHeaders);
-      tasks = normalizeJambones(this.logger, json).map((tdata) => makeTask(this.logger, tdata));
-    }
-    else if (Array.isArray(whisper)) {
+    // Disable passing a URL as not functional Sam Machin - 17/07/2025
+    //if (typeof whisper === 'string' || (typeof whisper === 'object' && whisper.url && !whisper.verb)) {
+    //  // retrieve a url
+    //  const json = await this.requestor(opts.call_hook, this.callInfo.toJSON(), httpHeaders);
+    //  tasks = normalizeJambones(this.logger, json).map((tdata) => makeTask(this.logger, tdata));
+    //}
+    //else if (Array.isArray(whisper)) {
+
+    if (Array.isArray(whisper)) {
       // an inline array of tasks
       tasks = normalizeJambones(this.logger, whisper).map((tdata) => makeTask(this.logger, tdata));
     }
@@ -1913,6 +1968,17 @@ Duration=${duration} `
       }
     } catch (err) {
       this.logger.info({err, opts, callSid}, 'CallSession:updateCall - error updating call');
+      const {writeAlerts} = this.srf.locals;
+      try {
+        writeAlerts({
+          alert_type: 'error-updating-call',
+          account_sid: this.accountSid,
+          message: err.message,
+          target_sid: callSid
+        });
+      } catch (err) {
+        this.logger.error({err}, 'Error writing error-updating-call alert');
+      }
     }
   }
 
@@ -2317,6 +2383,7 @@ Duration=${duration} `
       const ep = await this._createMediaEndpoint({
         headers: {
           'X-Jambones-Call-ID': this.callId,
+          'X-Call-Sid': this.callSid,
         },
         remoteSdp: this.req.body
       });
@@ -2461,6 +2528,8 @@ Duration=${duration} `
     this.clearOrRestoreActionHookDelayProcessor().catch((err) => {});
 
     this.ttsStreamingBuffer?.stop();
+
+    this.sttLatencyCalculator?.stop();
   }
 
   /**
@@ -3009,6 +3078,11 @@ Duration=${duration} `
     const task = this.currentTask;
     if (task && TaskName.Say === task.name) {
       task.notifyTtsStreamIsEmpty();
+    } else if (
+      // If Gather nested say task is streaming
+      task && TaskName.Gather === task.name && task.sayTask && task.sayTask.isStreamingTts) {
+      const sayTask = task.sayTask;
+      sayTask.notifyTtsStreamIsEmpty();
     }
   }
 
@@ -3103,6 +3177,20 @@ Duration=${duration} `
       return `assistant: ${t.text}`;
     }).join('\n');
   }
+
+  startSttLatencyVad() {
+    if (this.notifySttLatencyEnabled) {
+      this.sttLatencyCalculator.start();
+    }
+  }
+
+  stopSttLatencyVad() {
+    this.sttLatencyCalculator.stop();
+  }
+
+  calculateSttLatency() {
+    return this.sttLatencyCalculator.calculateLatency();
+  }
 }
 
 module.exports = CallSession;

lib/session/confirm-call-session.js

@@ -8,7 +8,8 @@ const CallSession = require('./call-session');
 
  */
 class ConfirmCallSession extends CallSession {
-  constructor({logger, application, dlg, ep, tasks, callInfo, accountInfo, memberId, confName, rootSpan, req}) {
+  // eslint-disable-next-line max-len
+  constructor({logger, application, dlg, ep, tasks, callInfo, accountInfo, memberId, confName, rootSpan, req, tmpFiles}) {
     super({
       logger,
       application,
@@ -24,6 +25,7 @@ class ConfirmCallSession extends CallSession {
     this.dlg = dlg;
     this.ep = ep;
     this.req = req;
+    this.tmpFiles = tmpFiles;
   }
 
   /**

lib/tasks/conference.js

@@ -673,7 +673,8 @@ class Conference extends Task {
         confName: this.confName,
         tasks,
         rootSpan: cs.rootSpan,
-        req: cs.req
+        req: cs.req,
+        tmpFiles: cs.tmpFiles,
       });
       await this._playSession.exec();
       this._playSession = null;

lib/tasks/config.js

@@ -24,6 +24,9 @@ class TaskConfig extends Task {
     if ('notifyEvents' in this.data) {
       this.notifyEvents = !!this.data.notifyEvents;
     }
+    if (this.hasNotifySttLatency) {
+      this.notifySttLatency = !!this.data.notifySttLatency;
+    }
 
     if (this.bargeIn.enable) {
       this.gatherOpts = {
@@ -83,6 +86,7 @@ class TaskConfig extends Task {
   get hasVad() { return Object.keys(this.vad).length; }
   get hasFillerNoise() { return Object.keys(this.fillerNoise).length; }
   get hasReferHook() { return Object.keys(this.data).includes('referHook'); }
+  get hasNotifySttLatency() { return Object.keys(this.data).includes('notifySttLatency'); }
   get hasTtsStream() { return Object.keys(this.ttsStream).length; }
 
   get summary() {
@@ -112,6 +116,8 @@ class TaskConfig extends Task {
     if (this.hasFillerNoise) phrase.push(`fillerNoise ${this.fillerNoise.enable ? 'on' : 'off'}`);
     if (this.data.amd) phrase.push('enable amd');
     if (this.notifyEvents) phrase.push(`event notification ${this.notifyEvents ? 'on' : 'off'}`);
+    if (this.hasNotifySttLatency) phrase.push(
+      `notifySttLatency ${this.notifySttLatency ? 'on' : 'off'}`);
     if (this.onHoldMusic) phrase.push(`onHoldMusic: ${this.onHoldMusic}`);
     if ('boostAudioSignal' in this.data) phrase.push(`setGain ${this.data.boostAudioSignal}`);
     if (this.hasReferHook) phrase.push('set referHook');
@@ -130,6 +136,11 @@ class TaskConfig extends Task {
       cs.notifyEvents = !!this.data.notifyEvents;
     }
 
+    if (this.hasNotifySttLatency) {
+      this.logger.debug(`turning notifySttLatency ${this.notifySttLatency ? 'on' : 'off'}`);
+      cs.notifySttLatencyEnabled = this.notifySttLatency;
+    }
+
     if (this.onHoldMusic) {
       cs.onHoldMusic = this.onHoldMusic;
     }
@@ -318,7 +329,10 @@ class TaskConfig extends Task {
         voiceMs: this.vad.voiceMs || 250,
         silenceMs: this.vad.silenceMs || 150,
         strategy: this.vad.strategy || 'one-shot',
-        mode: (this.vad.mode !== undefined && this.vad.mode !== null) ? this.vad.mode : 2
+        mode: (this.vad.mode !== undefined && this.vad.mode !== null) ? this.vad.mode : 2,
+        vendor: this.vad.vendor || 'silero',
+        threshold: this.vad.threshold || 0.5,
+        speechPadMs: this.vad.speechPadMs || 30,
       };
     }
 

lib/tasks/dial.js

@@ -19,7 +19,7 @@ const parseDecibels = require('../utils/parse-decibels');
 const debug = require('debug')('jambonz:feature-server');
 const {parseUri} = require('drachtio-srf');
 const {ANCHOR_MEDIA_ALWAYS,
-  JAMBONZ_DISABLE_DIAL_PAI_HEADER,
+  JAMBONZ_DIAL_PAI_HEADER,
   JAMBONES_DIAL_SBC_FOR_REGISTERED_USER} = require('../config');
 const { isOnhold, isOpusFirst } = require('../utils/sdp-utils');
 const { normalizeJambones } = require('@jambonz/verb-specifications');
@@ -109,6 +109,7 @@ class TaskDial extends Task {
     this.tag = this.data.tag;
     this.boostAudioSignal = this.data.boostAudioSignal;
     this._mediaPath = MediaPath.FullMedia;
+    this.forwardPAI = this.data.forwardPAI;
 
     if (this.dtmfHook) {
       const {parentDtmfCollector, childDtmfCollector} = parseDtmfOptions(logger, this.data.dtmfCapture || {});
@@ -548,13 +549,14 @@ class TaskDial extends Task {
     let sbcAddress = this.proxy || getSBC();
     const teamsInfo = {};
     let fqdn;
-
+    const forwardPAI = this.forwardPAI ?? JAMBONZ_DIAL_PAI_HEADER; // dial verb overides env var
+    this.logger.debug(forwardPAI, 'forwardPAI value');
     if (!sbcAddress) throw new Error('no SBC found for outbound call');
     this.headers = {
       'X-Account-Sid': cs.accountSid,
       ...(req && req.has('X-CID') && {'X-CID': req.get('X-CID')}),
       ...(direction === 'outbound' && callInfo.sbcCallid && {'X-CID': callInfo.sbcCallid}),
-      ...(!JAMBONZ_DISABLE_DIAL_PAI_HEADER && req && {
+      ...(req && forwardPAI && {
         ...(req.has('P-Asserted-Identity') && {'P-Asserted-Identity': req.get('P-Asserted-Identity')}),
         ...(req.has('Privacy') && {'Privacy': req.get('Privacy')}),
       }),
@@ -995,7 +997,7 @@ class TaskDial extends Task {
       this.epOther = null;
       this._mediaPath = releaseEntirely ? MediaPath.NoMedia : MediaPath.PartialMedia;
       this.logger.info(
-        `Dial:_releaseMedia - successfully released media from freewitch, media path is now ${this._mediaPath}`);
+        `Dial:_releaseMedia - successfully released media from freeswitch, media path is now ${this._mediaPath}`);
     } catch (err) {
       this.logger.info({err}, 'Dial:_releaseMedia error');
     }
@@ -1004,7 +1006,7 @@ class TaskDial extends Task {
   async reAnchorMedia(cs, sd) {
     if (cs.ep && sd.ep) return;
 
-    this.logger.info('Dial:reAnchorMedia - re-anchoring media to freewitch');
+    this.logger.info('Dial:reAnchorMedia - re-anchoring media to freeswitch');
     await Promise.all([sd.reAnchorMedia(this._mediaPath), cs.reAnchorMedia(this._mediaPath)]);
     this.epOther = cs.ep;
 
@@ -1012,7 +1014,7 @@ class TaskDial extends Task {
 
     this._mediaPath = MediaPath.FullMedia;
     this.logger.info(
-      `Dial:_releaseMedia - successfully re-anchored media to freewitch, media path is now ${this._mediaPath}`);
+      `Dial:_releaseMedia - successfully re-anchored media to freeswitch, media path is now ${this._mediaPath}`);
   }
 
   // Handle RE-INVITE hold from caller leg.
@@ -1096,7 +1098,8 @@ class TaskDial extends Task {
               accountInfo: this.cs.accountInfo,
               tasks,
               rootSpan: this.cs.rootSpan,
-              req: this.cs.req
+              req: this.cs.req,
+              tmpFiles: this.cs.tmpFiles,
             });
             await this._onHoldSession.exec();
             this._onHoldSession = null;

lib/tasks/enqueue.js

@@ -370,7 +370,8 @@ class TaskEnqueue extends Task {
         accountInfo: cs.accountInfo,
         tasks: tasksToRun,
         rootSpan: cs.rootSpan,
-        req: cs.req
+        req: cs.req,
+        tmpFiles: cs.tmpFiles,
       });
       await this._playSession.exec();
       this._playSession = null;

lib/tasks/gather.js

@@ -11,6 +11,7 @@ const {
   NvidiaTranscriptionEvents,
   JambonzTranscriptionEvents,
   AssemblyAiTranscriptionEvents,
+  DeepgramRiverTranscriptionEvents,
   VoxistTranscriptionEvents,
   CartesiaTranscriptionEvents,
   OpenAITranscriptionEvents,
@@ -267,11 +268,16 @@ class TaskGather extends SttTask {
           .catch((err) => {
             process();
           });
-        this.sayTask.on('playDone', (err) => {
-          span.end();
-          if (err) this.logger.error({err}, 'Gather:exec Error playing tts');
+        if (this.sayTask.isStreamingTts && !this.sayTask.closeOnStreamEmpty) {
+          // if streaming tts, we do not wait for it to complete if it is not closing the stream automatically
           process();
-        });
+        } else {
+          this.sayTask.on('playDone', (err) => {
+            span.end();
+            if (err) this.logger.error({err}, 'Gather:exec Error playing tts');
+            process();
+          });
+        }
       }
       else if (this.playTask) {
         const {span, ctx} = this.startChildSpan(`nested:${this.playTask.summary}`);
@@ -346,6 +352,7 @@ class TaskGather extends SttTask {
     this.sayTask?.span.end();
     this._stopVad();
     this._resolve('killed');
+    cs.stopSttLatencyVad();
   }
 
   updateTaskInProgress(opts) {
@@ -361,6 +368,9 @@ class TaskGather extends SttTask {
 
   _onDtmf(cs, ep, evt) {
     this.logger.debug(evt, 'TaskGather:_onDtmf');
+    if (!this._timeoutTimer && this.timeout > 0) {
+      this._startTimer();
+    }
     clearTimeout(this.interDigitTimer);
     let resolved = false;
     if (this.dtmfBargein) {
@@ -385,12 +395,7 @@ class TaskGather extends SttTask {
       if (this.digitBuffer.length === 0 && this.needsStt) {
         // DTMF is higher priority than STT.
         this.removeCustomEventListeners();
-        ep.stopTranscription({
-          vendor: this.vendor,
-          bugname: this.bugname,
-        })
-          .catch((err) => this.logger.error({err},
-            ` Received DTMF, Error stopping transcription for vendor ${this.vendor}`));
+        this._stopTranscribing(ep);
       }
       this.digitBuffer += evt.dtmf;
       const len = this.digitBuffer.length;
@@ -463,6 +468,16 @@ class TaskGather extends SttTask {
           this._onVendorConnectFailure.bind(this, cs, ep));
         break;
 
+      case 'deepgramriver':
+        this.bugname = `${this.bugname_prefix}deepgramriver_transcribe`;
+        this.addCustomEventListener(
+          ep, DeepgramRiverTranscriptionEvents.Transcription, this._onTranscription.bind(this, cs, ep));
+        this.addCustomEventListener(
+          ep, DeepgramRiverTranscriptionEvents.Connect, this._onVendorConnect.bind(this, cs, ep));
+        this.addCustomEventListener(ep, DeepgramRiverTranscriptionEvents.ConnectFailure,
+          this._onVendorConnectFailure.bind(this, cs, ep));
+        break;
+
       case 'soniox':
         this.bugname = `${this.bugname_prefix}soniox_transcribe`;
         this.addCustomEventListener(
@@ -677,10 +692,14 @@ class TaskGather extends SttTask {
         target_sid: this.cs.callSid
       });
     }).catch((err) => this.logger.info({err}, 'Error generating alert for tts failure'));
+
+    // Some vendor use single connection, that we cannot use onConnect event to track transcription start
+    this.cs.emit('transcribe-start');
   }
 
   _startTimer() {
     if (0 === this.timeout) return;
+    this.logger.debug(`Starting timoutTimer of ${this.timeout}ms`);
     this._clearTimer();
     this._timeoutTimer = setTimeout(() => {
       if (this.isContinuousAsr) this._startAsrTimer();
@@ -860,6 +879,10 @@ class TaskGather extends SttTask {
     if (finished === 'true') return;
 
     if (this.vendor === 'ibm' && evt?.state === 'listening') return;
+
+    // emit an event to the call session to track the time transcription is received
+    cs.emit('on-transcription');
+
     if (this.vendor === 'deepgram' && evt.type === 'UtteranceEnd') {
       /* we will only get this when we have set utterance_end_ms */
       if (this._bufferedTranscripts.length === 0) {
@@ -943,6 +966,12 @@ class TaskGather extends SttTask {
         }
       }
 
+      // receive a final transcript, calculate the stt latency for this transcript
+      const sttLatency = this.cs.calculateSttLatency();
+      if (!emptyTranscript && sttLatency) {
+        this.stt_latency_ms += `${sttLatency.stt_latency_ms},`;
+      }
+
       if (this.isContinuousAsr) {
         /* append the transcript and start listening again for asrTimeout */
         const t = evt.alternatives[0].transcript;
@@ -1094,12 +1123,7 @@ class TaskGather extends SttTask {
 
   async _startFallback(cs, ep, evt) {
     if (this.canFallback) {
-      ep.stopTranscription({
-        vendor: this.vendor,
-        bugname: this.bugname,
-        gracefulShutdown: false
-      })
-        .catch((err) => this.logger.error({err}, `Error stopping transcription for primary vendor ${this.vendor}`));
+      this._stopTranscribing(ep);
       try {
         this.logger.debug('gather:_startFallback');
         this.notifyError({ msg: 'ASR error',
@@ -1228,17 +1252,26 @@ class TaskGather extends SttTask {
     }
   }
 
+  async _stopTranscribing(ep) {
+    // Fix for https://github.com/jambonz/jambonz-feature-server/issues/1281
+    // We should immediately call stop transcription from gather
+    // so that next gather can start transcription immediately
+    ep.stopTranscription({
+      vendor: this.vendor,
+      bugname: this.bugname,
+      gracefulShutdown: false
+    })
+      .catch((err) => {
+        if (this.resolved) return;
+        this.logger.error({err}, 'Error stopping transcription');
+      });
+    this.cs.emit('transcribe-stop');
+  }
+
   async _resolve(reason, evt) {
     this.logger.info({evt}, `TaskGather:resolve with reason ${reason}`);
     if (this.needsStt && this.ep && this.ep.connected) {
-      this.ep.stopTranscription({
-        vendor: this.vendor,
-        bugname: this.bugname
-      })
-        .catch((err) => {
-          if (this.resolved) return;
-          this.logger.error({err}, 'Error stopping transcription');
-        });
+      this._stopTranscribing(this.ep);
     }
     if (this.resolved) {
       this.logger.debug('TaskGather:_resolve - already resolved');
@@ -1257,11 +1290,28 @@ class TaskGather extends SttTask {
     this._clearAsrTimer();
     this._clearFinalAsrTimer();
 
+    let sttLatencyMetrics = {};
+    if (this.needsStt) {
+      const sttLatency = this.cs.calculateSttLatency();
+      if (sttLatency) {
+        this.stt_latency_ms = this.stt_latency_ms.endsWith(',') ?
+          this.stt_latency_ms.slice(0, -1) : this.stt_latency_ms;
+        sttLatencyMetrics = {
+          'stt.latency_ms': this.stt_latency_ms,
+          'stt.talkspurts': JSON.stringify(sttLatency.talkspurts),
+          'stt.start_time': sttLatency.stt_start_time,
+          'stt.stop_time': sttLatency.stt_stop_time,
+          'stt.usage': sttLatency.stt_usage,
+        };
+      }
+    }
+
     this.span.setAttributes({
       channel: 1,
       'stt.label': this.label || 'None',
       'stt.resolve': reason,
-      'stt.result': JSON.stringify(evt)
+      'stt.result': JSON.stringify(evt),
+      ...sttLatencyMetrics
     });
 
     if (this.callSession && this.callSession.callGone) {
@@ -1289,6 +1339,9 @@ class TaskGather extends SttTask {
 
     let returnedVerbs = false;
     try {
+      const latencies = Object.fromEntries(
+        Object.entries(sttLatencyMetrics).map(([key, value]) => [key.replace('stt.', 'stt_'), value])
+      );
       if (reason.startsWith('dtmf')) {
         if (this.parentTask) this.parentTask.emit('dtmf', evt);
         else {
@@ -1302,7 +1355,7 @@ class TaskGather extends SttTask {
         else {
           this.emit('transcription', evt);
           this.logger.debug('TaskGather:_resolve - invoking performAction');
-          returnedVerbs = await this.performAction({speech: evt, reason: 'speechDetected'});
+          returnedVerbs = await this.performAction({speech: evt, reason: 'speechDetected', ...latencies});
           this.logger.debug({returnedVerbs}, 'TaskGather:_resolve - back from performAction');
         }
       }
@@ -1310,20 +1363,20 @@ class TaskGather extends SttTask {
         if (this.parentTask) this.parentTask.emit('timeout', evt);
         else {
           this.emit('timeout', evt);
-          returnedVerbs = await this.performAction({reason: 'timeout'});
+          returnedVerbs = await this.performAction({reason: 'timeout', ...latencies});
         }
       }
       else if (reason.startsWith('stt-error')) {
         if (this.parentTask) this.parentTask.emit('stt-error', evt);
         else {
           this.emit('stt-error', evt);
-          returnedVerbs = await this.performAction({reason: 'error', details: evt.error});
+          returnedVerbs = await this.performAction({reason: 'error', details: evt.error, ...latencies});
         }
       } else if (reason.startsWith('stt-low-confidence')) {
         if (this.parentTask) this.parentTask.emit('stt-low-confidence', evt);
         else {
           this.emit('stt-low-confidence', evt);
-          returnedVerbs = await this.performAction({speech:evt, reason: 'stt-low-confidence'});
+          returnedVerbs = await this.performAction({speech:evt, reason: 'stt-low-confidence', ...latencies});
         }
       }
     } catch (err) {  /*already logged error*/ }

lib/tasks/listen.js

@@ -1,5 +1,5 @@
 const Task = require('./task');
-const {TaskName, TaskPreconditions, ListenEvents, ListenStatus} = require('../utils/constants');
+const {TaskName, TaskPreconditions, ListenEvents, ListenStatus} = require('../utils/constants.json');
 const makeTask = require('./make_task');
 const moment = require('moment');
 const MAX_PLAY_AUDIO_QUEUE_SIZE = 10;
@@ -72,7 +72,7 @@ class TaskListen extends Task {
     } catch (err) {
       this.logger.info(err, `TaskListen:exec - error ${this.url}`);
     }
-    if (this.transcribeTask) this.transcribeTask.kill();
+    if (this.transcribeTask) this.transcribeTask.kill(cs);
     this._removeListeners(ep);
   }
 
@@ -103,9 +103,12 @@ class TaskListen extends Task {
     this.notifyTaskDone();
   }
 
-  async updateListen(status) {
+  async updateListen(status, silence = false) {
     if (!this.killed && this.ep && this.ep.connected) {
-      const args = this._bugname ? [this._bugname] : [];
+      const args = [
+        ...(this._bugname ? [this._bugname] : []),
+        ...(status === ListenStatus.Pause ? ([silence]) : []),
+      ];
       this.logger.info(`TaskListen:updateListen status ${status}`);
       switch (status) {
         case ListenStatus.Pause:

lib/tasks/llm/llms/ultravox_s2s.js

@@ -218,7 +218,7 @@ class TaskLlmUltravox_S2S extends Task {
   async _onServerEvent(_ep, evt) {
     let endConversation = false;
     const type = evt.type;
-    this.logger.debug({evt}, 'TaskLlmUltravox_S2S:_onServerEvent');
+    //this.logger.debug({evt}, 'TaskLlmUltravox_S2S:_onServerEvent');
 
     /* server errors of some sort */
     if (type === 'error') {

lib/tasks/play.js

@@ -6,7 +6,9 @@ class TaskPlay extends Task {
     super(logger, opts);
     this.preconditions = TaskPreconditions.Endpoint;
 
-    this.url = this.data.url;
+    this.url = this.data.url.includes('?')
+      ? this.data.url.split('?')[0] + '?' + this.data.url.split('?')[1].replaceAll('.', '%2E')
+      : this.data.url;
     this.seekOffset = this.data.seekOffset || -1;
     this.timeoutSecs = this.data.timeoutSecs || -1;
     this.loop = this.data.loop || 1;

lib/tasks/say.js

@@ -5,6 +5,28 @@ const pollySSMLSplit = require('polly-ssml-split');
 const { SpeechCredentialError } = require('../utils/error');
 const { sleepFor } = require('../utils/helpers');
 
+/**
+ * Discard unmatching responses:
+ * (1) I sent a playback id but get a response with a different playback id
+ * (2) I sent a playback id but get a response with no playback id
+ * (3) I did not send a playback id but get a response with a playback id
+ * (4) I sent a cache file but get a response with a different cache file
+ */
+
+const isMatchingEvent = (logger, filename, playbackId, evt) => {
+
+  if (!!playbackId && !!evt.variable_tts_playback_id && evt.variable_tts_playback_id === playbackId) {
+    //logger.debug({filename, playbackId, evt}, 'Say:isMatchingEvent - playbackId matched');
+    return true;
+  }
+  if (!!filename && !!evt.file && evt.file === filename) {
+    //logger.debug({filename, playbackId, evt}, 'Say:isMatchingEvent - filename matched');
+    return true;
+  }
+  logger.info({filename, playbackId, evt}, 'Say:isMatchingEvent - no match');
+  return false;
+};
+
 const breakLengthyTextIfNeeded = (logger, text) => {
   // As The text can be used for tts streaming, we need to break lengthy text into smaller chunks
 // HIGH_WATER_BUFFER_SIZE defined in tts-streaming-buffer.js
@@ -259,40 +281,32 @@ class TaskSay extends TtsTask {
     while (!this.killed && (this.loop === 'forever' || this.loop--) && ep?.connected) {
       let segment = 0;
       while (!this.killed && segment < filepath.length) {
+        const filename = filepath[segment];
         if (cs.isInConference) {
           const {memberId, confName, confUuid} = cs;
-          await this.playToConfMember(ep, memberId, confName, confUuid, filepath[segment]);
+          await this.playToConfMember(ep, memberId, confName, confUuid, filename);
         }
         else {
-          let playbackId;
-          const isStreaming = filepath[segment].startsWith('say:{');
+          const isStreaming = filename.startsWith('say:{');
           if (isStreaming) {
-            const arr = /^say:\{.*\}\s*(.*)$/.exec(filepath[segment]);
-            if (arr) this.logger.debug(`Say:exec sending streaming tts request: ${arr[1].substring(0, 64)}..`);
-          }
-          else {
-            this.logger.debug(`Say:exec sending ${filepath[segment].substring(0, 64)}`);
+            const arr = /^say:\{.*\}\s*(.*)$/.exec(filename);
+            if (arr) this.logger.debug(`Say:exec sending streaming tts request ${arr[1].substring(0, 64)}..`);
+            else this.logger.debug(`Say:exec sending ${filename.substring(0, 64)}`);
           }
 
           const onPlaybackStop = (evt) => {
             try {
-              this.logger.debug({evt},
-                `Say got playback-stop ${evt.variable_tts_playback_id ? evt.variable_tts_playback_id : ''}`);
-
-              /**
-               * If we got a playback id on both the start and stop events, and they don't match,
-               * then we must have received a playback-stop event for an earlier play request.
-               */
-              const unmatchedResponse = (!!playbackId && !!evt.variable_tts_playback_id) &&
-                evt.variable_tts_playback_id !== playbackId;
-              if (unmatchedResponse) {
-                this.logger.info({currentPlaybackId: playbackId, stopPPlaybackId: evt.variable_tts_playback_id},
+              const playbackId = this.getPlaybackId(segment);
+              const isMatch = isMatchingEvent(this.logger, filename, playbackId, evt);
+              if (!isMatch) {
+                this.logger.info({currentPlaybackId: playbackId, stopPlaybackId: evt.variable_tts_playback_id},
                   'Say:exec discarding playback-stop for earlier play');
                 ep.once('playback-stop',  this._boundOnPlaybackStop);
 
                 return;
               }
-
+              this.logger.debug({evt},
+                `Say got playback-stop ${evt.variable_tts_playback_id ? evt.variable_tts_playback_id : ''}`);
               this.notifyStatus({event: 'stop-playback'});
               this.notifiedPlayBackStop = true;
               const tts_error = evt.variable_tts_error;
@@ -331,6 +345,7 @@ class TaskSay extends TtsTask {
                 !this.disableTtsCache
               ) {
                 const text = parseTextFromSayString(this.text[segment]);
+                this.logger.debug({text, cacheFile: evt.variable_tts_cache_filename}, 'Say:exec cache tts');
                 addFileToCache(evt.variable_tts_cache_filename, {
                   account_sid,
                   vendor,
@@ -358,9 +373,17 @@ class TaskSay extends TtsTask {
           };
           this._boundOnPlaybackStop = onPlaybackStop.bind(this);
 
-          ep.once('playback-start', (evt) => {
+          const onPlaybackStart = (evt) => {
             try {
-              playbackId = evt.variable_tts_playback_id;
+              const playbackId = this.getPlaybackId(segment);
+              const isMatch = isMatchingEvent(this.logger, filename, playbackId, evt);
+              if (!isMatch) {
+                this.logger.info({currentPlaybackId: playbackId, startPlaybackId: evt.variable_tts_playback_id},
+                  'Say:exec playback-start - unmatched playback_id');
+                ep.once('playback-start',  this._boundOnPlaybackStart);
+                return;
+              }
+              ep.once('playback-stop',  this._boundOnPlaybackStop);
               this.logger.debug({evt},
                 `Say got playback-start ${evt.variable_tts_playback_id ? evt.variable_tts_playback_id : ''}`);
               if (this.otelSpan) {
@@ -374,15 +397,17 @@ class TaskSay extends TtsTask {
             } catch (err) {
               this.logger.info({err}, 'Error handling playback-start event');
             }
-          });
-          ep.once('playback-stop',  this._boundOnPlaybackStop);
+          };
+          this._boundOnPlaybackStart = onPlaybackStart.bind(this);
+
+          ep.once('playback-start',  this._boundOnPlaybackStart);
 
           // wait for playback-stop event received to confirm if the playback is successful
           this._playPromise = new Promise((resolve, reject) => {
             this._playResolve = resolve;
             this._playReject = reject;
           });
-          const r = await ep.play(filepath[segment]);
+          const r = await ep.play(filename);
           this.logger.debug({r}, 'Say:exec play result');
           try {
             // wait for playback-stop event received to confirm if the playback is successful
@@ -400,12 +425,12 @@ class TaskSay extends TtsTask {
             this._playResolve = null;
             this._playReject = null;
           }
-          if (filepath[segment].startsWith('say:{')) {
-            const arr = /^say:\{.*\}\s*(.*)$/.exec(filepath[segment]);
+          if (filename.startsWith('say:{')) {
+            const arr = /^say:\{.*\}\s*(.*)$/.exec(filename);
             if (arr) this.logger.debug(`Say:exec complete playing streaming tts request: ${arr[1].substring(0, 64)}..`);
           } else {
             // This log will print spech credentials in say command for tts stream mode
-            this.logger.debug(`Say:exec completed play file ${filepath[segment]}`);
+            this.logger.debug(`Say:exec completed play file ${filename}`);
           }
         }
         segment++;
@@ -452,6 +477,7 @@ class TaskSay extends TtsTask {
           .replace('playht_', 'playht.')
           .replace('cartesia_', 'cartesia.')
           .replace('rimelabs_', 'rimelabs.')
+          .replace('resemble_', 'resemble.')
           .replace('inworld_', 'inworld.')
           .replace('verbio_', 'verbio.')
           .replace('elevenlabs_', 'elevenlabs.');
@@ -517,6 +543,9 @@ const spanMapping = {
   'rimelabs.name_lookup_time_ms': 'name_lookup_ms',
   'rimelabs.connect_time_ms': 'connect_ms',
   'rimelabs.final_response_time_ms': 'final_response_ms',
+  // Resemble
+  'resemble.connect_time_ms': 'connect_ms',
+  'resemble.final_response_time_ms': 'final_response_ms',
   // inworld
   'inworld.name_lookup_time_ms': 'name_lookup_ms',
   'inworld.connect_time_ms': 'connect_ms',

lib/tasks/sip_refer.js

@@ -111,7 +111,12 @@ class TaskSipRefer extends Task {
     /* get IP address of the SBC to use as hostname if needed */
     const {host} = parseUri(dlg.remote.uri);
 
-    if (!referTo.startsWith('<') && !referTo.startsWith('sip') && !referTo.startsWith('"')) {
+    if (
+      !referTo.startsWith('<') &&
+      !referTo.startsWith('sip:') &&
+      !referTo.startsWith('"') &&
+      !referTo.startsWith('tel:')
+    ) {
       /* they may have only provided a phone number/user */
       referTo = `sip:${referTo}@${host}`;
     }
@@ -124,7 +129,12 @@ class TaskSipRefer extends Task {
     if (!referredByDisplayName) {
       referredByDisplayName = cs.req?.callingName;
     }
-    if (!referredBy.startsWith('<') && !referredBy.startsWith('sip') && !referredBy.startsWith('"')) {
+    if (
+      !referredBy.startsWith('<') &&
+      !referredBy.startsWith('sip:') &&
+      !referredBy.startsWith('"') &&
+      !referredBy.startsWith('tel:')
+    ) {
       /* they may have only provided a phone number/user */
       referredBy = `${referredByDisplayName ? `"${referredByDisplayName}"` : ''}<sip:${referredBy}@${host}>`;
     }

lib/tasks/stt-task.js

@@ -4,6 +4,7 @@ const crypto = require('crypto');
 const { TaskPreconditions, CobaltTranscriptionEvents } = require('../utils/constants');
 const { SpeechCredentialError } = require('../utils/error');
 const {JAMBONES_AWS_TRANSCRIBE_USE_GRPC} = require('../config');
+const {TaskName} = require('../utils/constants.json');
 
 /**
  * "Please insert turns here: {{turns:4}}"
@@ -84,6 +85,9 @@ class SttTask extends Task {
     /*bug name prefix */
     this.bugname_prefix = '';
 
+    // stt latency calculator
+    this.stt_latency_ms = '';
+
   }
 
   async exec(cs, {ep, ep2}) {
@@ -91,6 +95,12 @@ class SttTask extends Task {
     this.ep = ep;
     this.ep2 = ep2;
 
+    // start vad from stt latency calculator
+    if (this.name !== TaskName.Gather ||
+      this.name === TaskName.Gather && this.needsStt) {
+      cs.startSttLatencyVad();
+    }
+
     // use session preferences if we don't have specific verb-level settings.
     if (cs.recognizer) {
       for (const k in cs.recognizer) {
@@ -400,7 +410,7 @@ class SttTask extends Task {
     dgOptions.utteranceEndMs = dgOptions.utteranceEndMs || asrTimeout;
   }
 
-  _onVendorConnect(_cs, _ep) {
+  _onVendorConnect(cs, _ep) {
     this.logger.debug(`TaskGather:_on${this.vendor}Connect`);
   }
 

lib/tasks/transcribe.js

@@ -6,6 +6,7 @@ const {
   AwsTranscriptionEvents,
   AzureTranscriptionEvents,
   DeepgramTranscriptionEvents,
+  DeepgramRiverTranscriptionEvents,
   SonioxTranscriptionEvents,
   CobaltTranscriptionEvents,
   IbmTranscriptionEvents,
@@ -152,12 +153,15 @@ class TaskTranscribe extends SttTask {
         .catch((err) => this.logger.info(err, 'Error TaskTranscribe:kill'));
     }
 
+    this.cs.emit('transcribe-stop');
+
     return stopTranscription;
   }
 
   async kill(cs) {
     super.kill(cs);
     const stopTranscription = this._stopTranscription();
+    cs.stopSttLatencyVad();
     // hangup after 1 sec if we don't get a final transcription
     if (stopTranscription) this._timer = setTimeout(() => this.notifyTaskDone(), 1500);
     else this.notifyTaskDone();
@@ -238,6 +242,15 @@ class TaskTranscribe extends SttTask {
         //if (opts.DEEPGRAM_SPEECH_UTTERANCE_END_MS) this.isContinuousAsr = true;
 
         break;
+      case 'deepgramriver':
+        this.bugname = `${this.bugname_prefix}deepgramriver_transcribe`;
+        this.addCustomEventListener(ep, DeepgramRiverTranscriptionEvents.Transcription,
+          this._onTranscription.bind(this, cs, ep, channel));
+        this.addCustomEventListener(ep, DeepgramRiverTranscriptionEvents.Connect,
+          this._onVendorConnect.bind(this, cs, ep));
+        this.addCustomEventListener(ep, DeepgramRiverTranscriptionEvents.ConnectFailure,
+          this._onVendorConnectFailure.bind(this, cs, ep, channel));
+        break;
       case 'soniox':
         this.bugname = `${this.bugname_prefix}soniox_transcribe`;
         this.addCustomEventListener(ep, SonioxTranscriptionEvents.Transcription,
@@ -423,6 +436,9 @@ class TaskTranscribe extends SttTask {
       bugname: this.bugname,
       hostport: this.hostport
     });
+
+    // Some vendor use single connection, that we cannot use onConnect event to track transcription start
+    this.cs.emit('transcribe-start');
   }
 
   async _onTranscription(cs, ep, channel, evt, fsEvent) {
@@ -441,6 +457,9 @@ class TaskTranscribe extends SttTask {
 
     if (this.vendor === 'ibm' && evt?.state === 'listening') return;
 
+    // emit an event to the call session to track the time transcription is received
+    cs.emit('on-transcription');
+
     if (this.vendor === 'deepgram' && evt.type === 'UtteranceEnd') {
       /* we will only get this when we have set utterance_end_ms */
 
@@ -602,14 +621,28 @@ class TaskTranscribe extends SttTask {
   }
 
   async _resolve(channel, evt) {
+    let sttLatencyMetrics = {};
     if (evt.is_final) {
+      const sttLatency = this.cs.calculateSttLatency();
+      if (sttLatency) {
+        sttLatencyMetrics = {
+          'stt.latency_ms': `${sttLatency.stt_latency_ms}`,
+          'stt.talkspurts': JSON.stringify(sttLatency.talkspurts),
+          'stt.start_time': sttLatency.stt_start_time,
+          'stt.stop_time': sttLatency.stt_stop_time,
+          'stt.usage': sttLatency.stt_usage,
+        };
+      }
+      // time to reset the stt latency
+      this.cs.emit('transcribe-start');
       /* we've got a final transcript, so end the otel child span for this channel */
       if (this.childSpan[channel - 1] && this.childSpan[channel - 1].span) {
         this.childSpan[channel - 1].span.setAttributes({
           channel,
           'stt.label': this.label || 'None',
           'stt.resolve': 'transcript',
-          'stt.result': JSON.stringify(evt)
+          'stt.result': JSON.stringify(evt),
+          ...sttLatencyMetrics
         });
         this.childSpan[channel - 1].span.end();
       }
@@ -618,9 +651,13 @@ class TaskTranscribe extends SttTask {
     if (this.transcriptionHook) {
       const b3 = this.getTracingPropagation();
       const httpHeaders = b3 && {b3};
+      const latencies = Object.fromEntries(
+        Object.entries(sttLatencyMetrics).map(([key, value]) => [key.replace('stt.', 'stt_'), value])
+      );
       const payload = {
         ...this.cs.callInfo,
         ...httpHeaders,
+        ...latencies,
         ...(evt.alternatives && {speech: evt}),
         ...(evt.type && {speechEvent: evt})
       };

lib/tasks/tts-task.js

@@ -3,6 +3,16 @@ const { TaskPreconditions } = require('../utils/constants');
 const { SpeechCredentialError } = require('../utils/error');
 const dbUtils = require('../utils/db-utils');
 
+const extractPlaybackId = (str) => {
+  // Match say:{...} and capture the content inside braces
+  const match = str.match(/say:\{([^}]*)\}/);
+  if (!match) return null;
+
+  // Look for playback_id=value within the captured content
+  const playbackMatch = match[1].match(/playback_id=([^,]*)/);
+  return playbackMatch ? playbackMatch[1] : null;
+};
+
 class TtsTask extends Task {
 
   constructor(logger, data, parentTask) {
@@ -22,6 +32,11 @@ class TtsTask extends Task {
     this.disableTtsCache = this.data.disableTtsCache;
     this.options = this.synthesizer.options || {};
     this.instructions = this.data.instructions;
+    this.playbackIds = [];
+  }
+
+  getPlaybackId(offset) {
+    return this.playbackIds[offset];
   }
 
   async exec(cs) {
@@ -280,6 +295,7 @@ class TtsTask extends Task {
             renderForCaching: preCache
           });
           if (!filePath.startsWith('say:')) {
+            this.playbackIds.push(null);
             this.logger.debug(`Say: file ${filePath}, served from cache ${servedFromCache}`);
             if (filePath) cs.trackTmpFile(filePath);
             if (this.otelSpan) {
@@ -293,12 +309,24 @@ class TtsTask extends Task {
                 vendor,
                 language,
                 characters: text.length,
-                elapsedTime: rtt
+                elapsedTime: rtt,
+                servedFromCache,
+                'id': this.id
+              });
+            }
+            if (servedFromCache) {
+              this.notifyStatus({
+                event: 'synthesized-audio',
+                vendor,
+                language,
+                servedFromCache,
+                'id': this.id
               });
             }
           }
           else {
-            this.logger.debug('Say: a streaming tts api will be used');
+            this.playbackIds.push(extractPlaybackId(filePath));
+            this.logger.debug({playbackIds: this.playbackIds}, 'Say: a streaming tts api will be used');
             const modifiedPath = filePath.replace('say:{', `say:{session-uuid=${ep.uuid},`);
             return modifiedPath;
           }

lib/utils/amd-utils.js

@@ -281,13 +281,17 @@ module.exports = (logger) => {
 
     /* set stt options */
     logger.info(`starting amd for vendor ${vendor} and language ${language}`);
-    const sttOpts = amd.setChannelVarsForStt({name: TaskName.Gather}, sttCredentials, language, {
-      vendor,
-      hints,
-      enhancedModel: true,
-      altLanguages: opts.recognizer?.altLanguages || [],
-      initialSpeechTimeoutMs: opts.resolveTimeoutMs,
-    });
+    /* if opts contains recognizer object use that config for stt, otherwise use defaults */
+    const rOpts = opts.recognizer ?
+      opts.recognizer :
+      {
+        vendor,
+        hints,
+        enhancedModel: true,
+        altLanguages: opts.recognizer?.altLanguages || [],
+        initialSpeechTimeoutMs: opts.resolveTimeoutMs,
+      };
+    const sttOpts = amd.setChannelVarsForStt({name: TaskName.Gather}, sttCredentials, language, rOpts);
 
     await ep.set(sttOpts).catch((err) => logger.info(err, 'Error setting channel variables'));
 

lib/utils/background-task-manager.js

@@ -65,7 +65,7 @@ class BackgroundTaskManager extends Emitter {
       this.logger.info(`stopping background task: ${type}`);
       task.removeAllListeners();
       task.span.end();
-      task.kill();
+      task.kill(this.cs);
       // Remove task from managed List
       this.tasks.delete(type);
     }

lib/utils/constants.json

@@ -96,6 +96,11 @@
     "ConnectFailure": "deepgram_transcribe::connect_failed",
     "Connect": "deepgram_transcribe::connect"
   },
+  "DeepgramRiverTranscriptionEvents": {
+    "Transcription": "deepgramriver_transcribe::transcription",
+    "ConnectFailure": "deepgramriver_transcribe::connect_failed",
+    "Connect": "deepgramriver_transcribe::connect"
+  },
   "SonioxTranscriptionEvents": {
     "Transcription": "soniox_transcribe::transcription",
     "Error": "soniox_transcribe::error"
@@ -177,6 +182,9 @@
   "VadDetection": {
     "Detection": "vad_detect:detection"
   },
+  "SileroVadDetection": {
+    "Detection": "vad_silero:detect"
+  },
   "ListenEvents": {
     "Connect": "mod_audio_fork::connect",
     "ConnectFailure": "mod_audio_fork::connect_failed",

lib/utils/db-utils.js

@@ -81,6 +81,10 @@ const speechMapper = (cred) => {
       obj.deepgram_tts_uri = o.deepgram_tts_uri;
       obj.deepgram_stt_use_tls = o.deepgram_stt_use_tls;
     }
+    else if ('deepgramriver' === obj.vendor) {
+      const o = JSON.parse(decrypt(credential));
+      obj.api_key = o.api_key;
+    }
     else if ('soniox' === obj.vendor) {
       const o = JSON.parse(decrypt(credential));
       obj.api_key = o.api_key;
@@ -120,6 +124,12 @@ const speechMapper = (cred) => {
       obj.model_id = o.model_id;
       obj.options = o.options;
     }
+    else if ('resemble' === obj.vendor) {
+      const o = JSON.parse(decrypt(credential));
+      obj.api_key = o.api_key;
+      obj.resemble_tts_use_tls = o.resemble_tts_use_tls;
+      obj.resemble_tts_uri = o.resemble_tts_uri;
+    }
     else if ('inworld' === obj.vendor) {
       const o = JSON.parse(decrypt(credential));
       obj.api_key = o.api_key;

lib/utils/http-requestor.js

@@ -41,7 +41,7 @@ class HttpRequestor extends BaseRequestor {
   constructor(logger, account_sid, hook, secret) {
     super(logger, account_sid, hook, secret);
 
-    this.method = hook.method || 'POST';
+    this.method = hook.method?.toUpperCase() || 'POST';
     this.authHeader = basicAuth(hook.username, hook.password);
     this.backoffMs = 500;
 
@@ -111,7 +111,7 @@ class HttpRequestor extends BaseRequestor {
 
     const payload = params ? snakeCaseKeys(params, ['customerData', 'sip', 'env_vars', 'args']) : null;
     const url = hook.url || hook;
-    const method = hook.method || 'POST';
+    const method = hook.method?.toUpperCase() || 'POST';
     let buf = '';
     httpHeaders = {
       ...httpHeaders,
@@ -119,7 +119,7 @@ class HttpRequestor extends BaseRequestor {
     };
 
     assert.ok(url, 'HttpRequestor:request url was not provided');
-    assert.ok, (['GET', 'POST'].includes(method), `HttpRequestor:request method must be 'GET' or 'POST' not ${method}`);
+    assert.ok(['GET', 'POST'].includes(method), `HttpRequestor:request method must be 'GET' or 'POST' not ${method}`);
     const startAt = process.hrtime();
 
     /* if we have an absolute url, and it is ws then do a websocket connection */

lib/utils/place-outdial.js

@@ -401,7 +401,8 @@ class SingleDialer extends Emitter {
         accountInfo: this.accountInfo,
         tasks,
         rootSpan: this.rootSpan,
-        req: this.req
+        req: this.req,
+        tmpFiles: cs.tmpFiles,
       });
       await cs.exec();
 

lib/utils/process-monitor.js

@@ -0,0 +1,91 @@
+// lib/utils/process-monitor.js
+const fs = require('fs');
+const path = require('path');
+
+class ProcessMonitor {
+  constructor(logger) {
+    this.logger = logger;
+    this.packageInfo = this.getPackageInfo();
+    this.processName = this.packageInfo.name || 'unknown-app';
+  }
+
+  getPackageInfo() {
+    try {
+      const packagePath = path.join(process.cwd(), 'package.json');
+      return JSON.parse(fs.readFileSync(packagePath, 'utf8'));
+    } catch (e) {
+      return { name: 'unknown', version: 'unknown' };
+    }
+  }
+
+  logStartup(additionalInfo = {}) {
+    const startupInfo = {
+      msg: `${this.processName} started`,
+      app_name: this.processName,
+      app_version: this.packageInfo.version,
+      pid: process.pid,
+      ppid: process.ppid,
+      pm2_instance_id: process.env.NODE_APP_INSTANCE || 'not_pm2',
+      pm2_id: process.env.pm_id,
+      is_pm2: !!process.env.PM2,
+      node_version: process.version,
+      uptime: process.uptime(),
+      timestamp: new Date().toISOString(),
+      ...additionalInfo
+    };
+
+    this.logger.info(startupInfo);
+    return startupInfo;
+  }
+
+  setupSignalHandlers() {
+    // Log when we receive signals that would cause restart
+    process.on('SIGINT', () => {
+      this.logger.info({
+        msg: 'SIGINT received',
+        app_name: this.processName,
+        pid: process.pid,
+        ppid: process.ppid,
+        uptime: process.uptime(),
+        timestamp: new Date().toISOString()
+      });
+      process.exit(0);
+    });
+
+    process.on('SIGTERM', () => {
+      this.logger.info({
+        msg: 'SIGTERM received',
+        app_name: this.processName,
+        pid: process.pid,
+        ppid: process.ppid,
+        uptime: process.uptime(),
+        timestamp: new Date().toISOString()
+      });
+      process.exit(0);
+    });
+
+    process.on('uncaughtException', (error) => {
+      this.logger.error({
+        msg: 'Uncaught exception - process will restart',
+        app_name: this.processName,
+        error: error.message,
+        stack: error.stack,
+        pid: process.pid,
+        timestamp: new Date().toISOString()
+      });
+      process.exit(1);
+    });
+
+    process.on('unhandledRejection', (reason, promise) => {
+      this.logger.error({
+        msg: 'Unhandled rejection',
+        app_name: this.processName,
+        reason,
+        pid: process.pid,
+        timestamp: new Date().toISOString()
+      });
+    });
+  }
+}
+
+module.exports = ProcessMonitor;

lib/utils/stt-latency-calculator.js

@@ -0,0 +1,197 @@
+const { assert } = require('console');
+const Emitter = require('events');
+const {
+  VadDetection,
+  SileroVadDetection
+} = require('../utils/constants.json');
+
+class SttLatencyCalculator extends Emitter {
+  constructor({ logger, cs}) {
+    super();
+    this.logger = logger;
+    this.cs = cs;
+    this.isRunning = false;
+    this.isInTalkSpurt = false;
+    this.start_talking_time = 0;
+    this.talkspurts = [];
+    this.vendor = this.cs.vad?.vendor || 'silero';
+    this.stt_start_time = 0;
+    this.stt_stop_time = 0;
+    this.stt_on_transcription_time = 0;
+  }
+
+  set sttStartTime(time) {
+    this.stt_start_time = time;
+  }
+
+  get sttStartTime() {
+    return this.stt_start_time || 0;
+  }
+
+  set sttStopTime(time) {
+    this.stt_stop_time = time;
+  }
+
+  get sttStopTime() {
+    return this.stt_stop_time || 0;
+  }
+
+  set sttOnTranscriptionTime(time) {
+    this.stt_on_transcription_time = time;
+  }
+
+  get sttOnTranscriptionTime() {
+    return this.stt_on_transcription_time || 0;
+  }
+
+  _onVadDetected(_ep, _evt, fsEvent) {
+    if (fsEvent.getHeader('detected-event') === 'stop_talking') {
+      if (this.isInTalkSpurt) {
+        this.talkspurts.push({
+          start: this.start_talking_time,
+          stop: Date.now()
+        });
+      }
+
+      this.start_talking_time = 0;
+      this.isInTalkSpurt = false;
+    } else if (fsEvent.getHeader('detected-event') === 'start_talking') {
+      this.start_talking_time = Date.now();
+      this.isInTalkSpurt = true;
+    }
+  }
+
+  _startVad() {
+    assert(!this.isRunning, 'Latency calculator is already running');
+    assert(this.cs.ep, 'Callsession has no endpoint to start the latency calculator');
+    const ep = this.cs.ep;
+    if (!ep.sttLatencyVadHandler) {
+      ep.sttLatencyVadHandler = this._onVadDetected.bind(this, ep);
+      if (this.vendor === 'silero') {
+        ep.addCustomEventListener(SileroVadDetection.Detection, ep.sttLatencyVadHandler);
+      } else {
+        ep.addCustomEventListener(VadDetection.Detection, ep.sttLatencyVadHandler);
+      }
+    }
+    this.stop_talking_time = 0;
+    this.start_talking_time = 0;
+    this.vad = {
+      ...(this.cs.vad || {}),
+      strategy: 'continuous',
+      bugname: 'stt-latency-calculator-vad',
+      vendor: this.vendor
+    };
+
+    ep.startVadDetection(this.vad);
+    this.isRunning = true;
+  }
+
+  _stopVad() {
+    if (this.isRunning) {
+      this.logger.warn('Latency calculator is still running, stopping VAD detection');
+      const ep = this.cs.ep;
+      ep.stopVadDetection(this.vad);
+      if (ep.sttLatencyVadHandler) {
+        if (this.vendor === 'silero') {
+          this.ep?.removeCustomEventListener(SileroVadDetection.Detection, ep.sttLatencyVadHandler);
+        } else {
+          this.ep?.removeCustomEventListener(VadDetection.Detection, ep.sttLatencyVadHandler);
+        }
+        ep.sttLatencyVadHandler = null;
+      }
+      this.isRunning = false;
+      this.logger.info('STT Latency Calculator stopped');
+    }
+  }
+
+  start() {
+    if (this.isRunning) {
+      this.logger.warn('Latency calculator is already running');
+      return;
+    }
+    if (!this.cs.ep) {
+      this.logger.error('Callsession has no endpoint to start the latency calculator');
+      return;
+    }
+    this._startVad();
+    this.logger.debug('STT Latency Calculator started');
+  }
+
+  stop() {
+    this._stopVad();
+  }
+
+  toUnixTimestamp(date) {
+    return Math.floor(date / 1000);
+  }
+
+  calculateLatency() {
+    if (!this.isRunning) {
+      this.logger.debug('Latency calculator is not running, cannot calculate latency, returning default values');
+      return null;
+    }
+
+    const stt_stop_time = this.stt_stop_time || Date.now();
+    if (this.isInTalkSpurt) {
+      this.talkspurts.push({
+        start: this.start_talking_time,
+        stop: stt_stop_time
+      });
+      this.isInTalkSpurt = false;
+      this.start_talking_time = 0;
+    }
+    const stt_on_transcription_time = this.stt_on_transcription_time || stt_stop_time;
+    const start_talking_time = this.talkspurts[0]?.start;
+    let lastIdx = this.talkspurts.length - 1;
+    lastIdx = lastIdx < 0 ? 0 : lastIdx;
+    const stop_talking_time = this.talkspurts[lastIdx]?.stop || stt_stop_time;
+
+    return {
+      stt_start_time: this.toUnixTimestamp(this.stt_start_time),
+      stt_stop_time: this.toUnixTimestamp(stt_stop_time),
+      start_talking_time: this.toUnixTimestamp(start_talking_time),
+      stop_talking_time: this.toUnixTimestamp(stop_talking_time),
+      stt_latency: parseFloat((Math.abs(stt_on_transcription_time - stop_talking_time)) / 1000).toFixed(2),
+      stt_latency_ms: Math.abs(stt_on_transcription_time - stop_talking_time),
+      stt_usage: parseFloat((stt_stop_time - this.stt_start_time) / 1000).toFixed(2),
+      talkspurts: this.talkspurts.map((ts) =>
+        ([this.toUnixTimestamp(ts.start || 0), this.toUnixTimestamp(ts.stop || 0)]))
+    };
+  }
+
+  resetTime() {
+    if (!this.isRunning) {
+      return;
+    }
+    this.stt_start_time = Date.now();
+    this.stt_stop_time = 0;
+    this.stt_on_transcription_time = 0;
+    this.clearTalkspurts();
+    this.logger.info('STT Latency Calculator reset');
+  }
+
+  onTranscriptionReceived() {
+    if (!this.isRunning) {
+      return;
+    }
+    this.stt_on_transcription_time = Date.now();
+    this.logger.debug(`CallSession:on-transcription set to ${this.stt_on_transcription_time}`);
+  }
+
+  onTranscribeStop() {
+    if (!this.isRunning) {
+      return;
+    }
+    this.stt_stop_time = Date.now();
+    this.logger.debug(`CallSession:transcribe-stop set to ${this.stt_stop_time}`);
+  }
+
+  clearTalkspurts() {
+    this.talkspurts = [];
+    if (!this.isInTalkSpurt) {
+      this.start_talking_time = 0;
+    }
+  }
+}
+
+module.exports = SttLatencyCalculator;

lib/utils/transcription-utils.js

@@ -339,6 +339,25 @@ const normalizeDeepgram = (evt, channel, language, shortUtterance) => {
   };
 };
 
+const normalizeDeepgramRiver = (evt, channel, language, shortUtterance) => {
+  const copy = JSON.parse(JSON.stringify(evt));
+  return {
+    language_code: language,
+    channel_tag: channel,
+    is_final: evt.event === 'EndOfTurn',
+    alternatives: [
+      {
+        confidence: evt.end_of_turn_confidence,
+        transcript: evt.transcript,
+      }
+    ],
+    vendor: {
+      name: 'deepgramriver',
+      evt: copy
+    }
+  };
+};
+
 const normalizeNvidia = (evt, channel, language) => {
   const copy = JSON.parse(JSON.stringify(evt));
   const alternatives = (evt.alternatives || [])
@@ -650,6 +669,8 @@ module.exports = (logger) => {
     switch (vendor) {
       case 'deepgram':
         return normalizeDeepgram(evt, channel, language, shortUtterance);
+      case 'deepgramriver':
+        return normalizeDeepgramRiver(evt, channel, language, shortUtterance);
       case 'microsoft':
         return normalizeMicrosoft(evt, channel, language, punctuation);
       case 'google':
@@ -764,12 +785,15 @@ module.exports = (logger) => {
           AWS_ACCESS_KEY_ID: sttCredentials.accessKeyId,
           AWS_SECRET_ACCESS_KEY: sttCredentials.secretAccessKey,
           AWS_REGION: sttCredentials.region,
-          AWS_SECURITY_TOKEN: sttCredentials.securityToken
+          AWS_SECURITY_TOKEN: sttCredentials.securityToken,
+          AWS_SESSION_TOKEN: sttCredentials.sessionToken ? sttCredentials.sessionToken : sttCredentials.securityToken
         }),
         ...(awsOptions.accessKey && {AWS_ACCESS_KEY_ID: awsOptions.accessKey}),
         ...(awsOptions.secretKey && {AWS_SECRET_ACCESS_KEY: awsOptions.secretKey}),
         ...(awsOptions.region && {AWS_REGION: awsOptions.region}),
         ...(awsOptions.securityToken && {AWS_SECURITY_TOKEN: awsOptions.securityToken}),
+        ...(awsOptions.sessionToken && {AWS_SESSION_TOKEN: awsOptions.sessionToken ?
+          awsOptions.sessionToken : awsOptions.securityToken}),
         ...(awsOptions.languageModelName && {AWS_LANGUAGE_MODEL_NAME: awsOptions.languageModelName}),
         ...(awsOptions.piiEntityTypes?.length && {AWS_PII_ENTITY_TYPES: awsOptions.piiEntityTypes.join(',')}),
         ...(awsOptions.piiIdentifyEntities && {AWS_PII_IDENTIFY_ENTITIES: true}),
@@ -870,6 +894,14 @@ module.exports = (logger) => {
       const deepgramUri = deepgramOptions.deepgramSttUri || sttCredentials.deepgram_stt_uri;
       const useTls = deepgramOptions.deepgramSttUseTls || sttCredentials.deepgram_stt_use_tls;
 
+      // DH (2025-08-11) entity_prompt is currently limited to 100 words
+      const entityPrompt = deepgramOptions.entityPrompt ?
+        deepgramOptions.entityPrompt
+          .split(/\s+/)
+          .slice(0, 100)
+          .join(' ')
+        : undefined;
+
       /* default to a sensible model if not supplied */
       if (!model) {
         model = selectDefaultDeepgramModel(task, language);
@@ -928,7 +960,24 @@ module.exports = (logger) => {
         ...(deepgramOptions.fillerWords) &&
           {DEEPGRAM_SPEECH_FILLER_WORDS: deepgramOptions.fillerWords},
         ...((Array.isArray(deepgramOptions.keyterms) && deepgramOptions.keyterms.length > 0) &&
-          {DEEPGRAM_SPEECH_KEYTERMS: deepgramOptions.keyterms.join(',')})
+          {DEEPGRAM_SPEECH_KEYTERMS: deepgramOptions.keyterms.join(',')}),
+        ...(deepgramOptions.mipOptOut && {DEEPGRAM_SPEECH_MIP_OPT_OUT: deepgramOptions.mipOptOut}),
+        ...(entityPrompt && {DEEPGRAM_SPEECH_ENTITY_PROMPT: entityPrompt}),
+      };
+    }
+    else if ('deepgramriver' === vendor) {
+      const {
+        preflightThreshold,
+        eotThreshold,
+        eotTimeoutMs,
+        mipOptOut
+      } = rOpts.deepgramOptions || {};
+      opts = {
+        DEEPGRAM_API_KEY: sttCredentials.api_key,
+        ...(preflightThreshold && {DEEPGRAM_SPEECH_PRELIGHT_THRESHOLD: preflightThreshold}),
+        ...(eotThreshold && {DEEPGRAM_SPEECH_EOT_THRESHOLD: eotThreshold}),
+        ...(eotTimeoutMs && {DEEPGRAM_SPEECH_EOT_TIMEOUT_MS: eotTimeoutMs}),
+        ...(mipOptOut && {DEEPGRAM_SPEECH_MIP_OPT_OUT: mipOptOut}),
       };
     }
     else if ('soniox' === vendor) {

lib/utils/ws-requestor.js

@@ -293,7 +293,7 @@ class WsRequestor extends BaseRequestor {
 
       /* send the message */
       this.ws.send(JSON.stringify(obj), async() => {
-        this.logger.debug({obj}, `WsRequestor:request websocket: sent (${url})`);
+        if (obj.type !== 'llm:event') this.logger.debug({obj}, `WsRequestor:request websocket: sent (${url})`);
         // If session:reconnect is waiting for ack, hold here until ack to send queuedMsgs
         if (this._reconnectPromise) {
           try {

package.json

@@ -27,14 +27,14 @@
   "dependencies": {
     "@aws-sdk/client-auto-scaling": "^3.549.0",
     "@aws-sdk/client-sns": "^3.549.0",
-    "@jambonz/db-helpers": "^0.9.12",
+    "@jambonz/db-helpers": "^0.9.16",
     "@jambonz/http-health-check": "^0.0.1",
     "@jambonz/mw-registrar": "^0.2.7",
-    "@jambonz/realtimedb-helpers": "^0.8.13",
-    "@jambonz/speech-utils": "^0.2.15",
+    "@jambonz/realtimedb-helpers": "^0.8.15",
+    "@jambonz/speech-utils": "^0.2.22",
     "@jambonz/stats-collector": "^0.1.10",
     "@jambonz/time-series": "^0.2.14",
-    "@jambonz/verb-specifications": "^0.0.108",
+    "@jambonz/verb-specifications": "^0.0.113",
     "@modelcontextprotocol/sdk": "^1.9.0",
     "@opentelemetry/api": "^1.8.0",
     "@opentelemetry/exporter-jaeger": "^1.23.0",
@@ -48,7 +48,7 @@
     "bent": "^7.3.12",
     "debug": "^4.3.4",
     "deepcopy": "^2.1.0",
-    "drachtio-fsmrf": "^4.0.4",
+    "drachtio-fsmrf": "^4.1.2",
     "drachtio-srf": "^5.0.5",
     "express": "^4.19.2",
     "express-validator": "^7.0.1",

test/sip-refer-tests.js

@@ -58,6 +58,46 @@ test('\'refer\' tests w/202 and NOTIFY', {timeout: 25000}, async(t) => {
   }
 });
 
+test('\'refer\' tests tel:', {timeout: 25000}, async(t) => {
+  clearModule.all();
+  const {srf, disconnect} = require('../app');
+
+  try {
+    await connect(srf);
+
+    // GIVEN
+    const verbs = [
+      {
+        verb: 'say',
+        text: 'silence_stream://100'
+      },
+      {
+        verb: 'sip:refer',
+        referTo: 'tel:+1234567890',
+        actionHook: '/actionHook'
+      }
+    ];
+    const noVerbs = [];
+
+    const from = 'refer_with_tel';
+    await provisionCallHook(from, verbs);
+    await provisionActionHook(from, noVerbs)
+
+    // THEN
+    await sippUac('uac-refer-with-notify.xml', '172.38.0.10', from);
+    t.pass('refer: successfully received 202 Accepted');
+    await sleepFor(1000);
+    const obj  = await getJSON(`http:127.0.0.1:3100/lastRequest/${from}_actionHook`);
+    t.ok(obj.body.final_referred_call_status === 200, 'refer: successfully received NOTIFY with 200 OK');
+    // console.log(`obj: ${JSON.stringify(obj)}`);
+    disconnect();
+  } catch (err) {
+    console.log(`error received: ${err}`);
+    disconnect();
+    t.error(err);
+  }
+});
+
 test('\'refer\' tests w/202 but no NOTIFY', {timeout: 25000}, async(t) => {
   clearModule.all();
   const {srf, disconnect} = require('../app');