01 AI-native revenue intelligence

Your calls already know who's going to buy.

CallInsight AI listens to every sales call your team makes — then transcribes, scores, flags and coaches automatically. Hot-lead verdicts with evidence quotes, compliance checks, and knowledge that builds itself. Minutes after hangup.

Plugs into Freshworks · Deepgram transcription · WhatsApp sync · REST API

HOT LEAD 8 / 9 CRITERIA · CONF 0.91
Call — Aylin K. ↔ Prospect
today 14:32 · 12m 41s · freshcaller · inbound
ANALYZING
sentiment positive · 0.92 financial readiness ✓ decision-ready ✓ flag · price concern summary · 4 next steps
Prospect · 00:08:12 “We've already compared three clinics — if the dates work for March, we're ready to put down the deposit this week.”
100%
of calls transcribed & analyzed — not a 2% QA sample
9-point
hot-lead verdict with evidence quotes per criterion
11
built-in flag types, from compliance to deal-at-risk
17
pipeline workers running while your team sleeps
02 The pipeline

From hangup to insight, on autopilot.

No uploads, no tagging, no listening back. The moment a call ends in your CRM, a 17-worker async pipeline takes over.

01

Sync

Calls and recordings pull automatically from Freshworks & Freshcaller — deals, agents and metadata included.

sync_calls · download_audio
02

Transcribe

Deepgram speech-to-text with speaker diarization, then AI-corrected speaker labels and transcript repair.

deepgram · gemini 2.5
03

Analyze

Sentiment, executive summary, compliance check, coaching opportunities and Q&A extraction — every call.

execute_ai_tasks
04

Score

The 9-criteria hot-lead evaluator delivers a verdict with confidence scores and transcript evidence.

call_evaluator
05

Flag

Your rules engine raises severity-ranked flags and routes them to the right owner in real time.

detect_flags
06

Learn

Questions cluster into FAQ candidates, knowledge gaps surface, and embeddings keep search semantic.

match_and_validate

avg. wall-clock from hangup → full analysis: minutes, not Mondays

03 Call intelligence

Every call becomes structured data.

Five analysis engines run on every conversation — built on Gemini 2.5, GPT-5 and Claude via a vendor-agnostic prompt registry you can extend.

Sentiment analysis

Customer and agent sentiment tracked separately, with emotional triggers and the exact moments the mood shifted.

overall · confidence · change_points · coaching_notes

Executive summary

A 3–5 sentence brief with key discussion points, the call outcome, and concrete next steps — readable in 20 seconds.

summary · key_points · outcome · next_steps

Compliance check

Recording consent, privacy disclosure, pricing transparency — verified on every call with a risk level and a list of what's missing.

compliance_score · risk_level · missing_elements

Coaching opportunities

Strengths and improvement areas per agent, with timestamped examples pulled straight from the transcript.

strengths · improvement_areas · examples[ts]

Q&A extraction

Every question a prospect asks — and how your agent answered — extracted, embedded and validated against your knowledge base.

question · answer · faq_match · verdict

Explore: ask anything

Natural-language questions over your entire call library, with saved analysis recipes you can version, pin and re-run.

semantic search · recipes · run history
Hot-lead scoring

Nine signals. One verdict. Zero gut feel.

The call evaluator scores every prospect against nine buying-intent criteria — from a concrete operation window to financial readiness to photo-consult agreement. Each criterion carries a confidence score, an ambiguity flag, and up to two verbatim evidence quotes from the transcript. Your team opens the morning dashboard and calls the right people first.

Lead verdict · call #4821 8/9 · HOT
Specific operation time window 0.94
Pre-research completed 0.89
Detailed questions asked 0.97
Financial readiness 0.88
Responsive & engaged 0.92
Decision-ready stage 0.85
Brand awareness 0.78
Not price-only focused ambiguous
Photos / consult ready 0.90
“I've sent my photos to your WhatsApp already — can we do the video consult on Thursday?” EVIDENCE · CRITERION 9 · 00:10:47
04 Monitoring & knowledge

Nothing critical slips through. Nothing learned gets lost.

Smart flagging

An alarm system for your revenue.

Eleven built-in flag types plus a custom rules engine — keyword, metadata, AI-result and composite AND/OR conditions. Flags carry severity, get auto-assigned, and live on the dashboard until someone resolves them.

CRITICALCompliance violationauto-assign → QA
HIGHDeal at risknotify · manager
HIGHUnresolved objectionfollow-up
MEDIUMCompetitor mentionedtagged
MEDIUMPrice concerntagged
INFOCommitment madelogged
Status workflow: open → acknowledged → resolved / dismissed · full audit trail
Self-building knowledge

Your FAQ writes itself from real calls.

Recurring questions cluster automatically into FAQ candidates with drafted answers. Agent answers are validated against your knowledge base with semantic matching + NLI — contradictions become violations, unknowns become knowledge gaps.

"Is the hotel included in the package price?"
asked 47× this monthFAQ matched · entails
"Can I pay the deposit in installments?"
asked 31×KNOWLEDGE GAP → candidate drafted
"How long before I can fly home?"
asked 26×FAQ matched · 0.91 sim
pgvector semantic search · auto-verdicts: ok / gap / violation / needs_review · human validation queue
05 Coaching

Train on Tuesday what the calls taught you on Monday.

An AI sparring partner for your agents — built from your real conversations, not generic scripts.

training session · "price objection — strong intensity" · persona: skeptical researcher
AI CustomerHonestly, the other clinic quoted me 30% less. Why would I pay more for the same procedure?
YouThat's a fair question — can I walk you through what's actually included in our package? The difference is usually in aftercare.
AI CustomerEveryone says that. What specifically is in the aftercare?
Live coaching
Good reframe — you acknowledged before countering. Now get specific: name 2 concrete aftercare items with outcomes.
Objection handling
7.8
Empathy
8.5
Value articulation
6.2
Confidence
7.4
Closing skills
5.8
Response quality
8.1
01

Roleplay real scenarios

Scenarios by difficulty with customizable personas and objection intensity — mild, moderate or strong. Locked progression keeps it a game agents want to win.

02

Coaching while they type

Instant feedback after every response, six live skill meters, and an objectives tracker. Or end-of-session mode for a full debrief with per-message scores.

03

Real-time assist on live calls

The live-assist module pushes sub-150-character suggestions to agents mid-call — objection counters and next moves in under 100ms, powered by Claude.

04

Scored like a real call

Session summaries use the same skill taxonomy as your production coaching analysis — so training scores and call scores speak the same language.

WhatsApp, unified

Per-agent WhatsApp sessions sync every message, media file and read receipt into the same customer record as the calls — automatically.

One customer, one timeline

Calls, WhatsApp threads, notes and files in one chronological view — with health scores, contact stats and deal context attached.

SMS, voice & email ready

Telnyx numbers for SMS and voice, Brevo for transactional email — every channel lands in the same timeline.

06 Platform

Built for teams. Open for builders.

Everything your analysts see in the dashboard, your systems can pull over a key-authenticated REST API.

Your data, one GET away.

Pull every call on a deal — transcripts, sentiment, summaries, Q&A and signed audio links — into your BI stack, your CRM automations, or anything that speaks HTTP.

Freshworks CRMauto call & deal sync
Freshcallernative call ingestion
Deepgramspeech-to-text + diarization
OpenRouterGemini · GPT-5 · Claude
WhatsApp (WAHA)per-agent message sync
TelnyxSMS & voice numbers
Brevotransactional email
S3 / MinIOyour bucket, your audio
api.callinsight.ai/api/deals/:deal_id GET · API KEY
{
  "deal_id": "31002048117",
  "calls": [{
    "date": "2026-06-09T14:32:11Z",
    "duration_seconds": 761,
    "transcript": [ … 142 utterances … ],
    "analysis": {
      "sentiment": { "overall": "positive", "confidence": 0.92 },
      "lead_verdict": { "score": "8/9", "label": "hot" },
      "summary": "Prospect compared 3 clinics…",
      "flags": [ "price_concern" ]
    },
    "audio_url": "…/api/audio/4821?token=sgn_…"
  }]
}
True tenant isolation

Schema-per-tenant PostgreSQL. Your data never shares a table with anyone else's — not row-level filters, separate schemas.

Real-time by construction

Phoenix LiveView dashboards update the second a worker finishes. No refresh button, no polling.

Signed audio access

Recordings stream through expiring, token-signed URLs from your own S3/MinIO bucket. Audio never leaks.

Human-in-the-loop QA

Validation queues for AI verdicts, FAQ candidates and answer reviews — approve, reject, or correct with full audit history.

Stop sampling calls. Start hearing all of them.

Connect your CRM in an afternoon. By tomorrow morning, every call your team made today is scored, flagged and searchable.