Jev call copilot for CurrencyTransfer

Working documents for a live call copilot built on Cloudflare Workers and TypeSafe's Jev model, modelled on the open-source sales copilot and adapted for onboarding and customer-success calls. Read the architecture first; the research documents are its evidence.

Documents

Architecture

Architecture14,072 words
The synthesised design: components on Cloudflare, data model, the per-utterance decision loop, curation, evaluation, compliance, and milestone 1.
Proposal A: fastest proof of concept6,524 words
Least machinery to get milestone 1 in front of you.
Proposal B: production-shaped6,153 words
The shape the system needs in production, with milestone 1 as its first vertical slice.
Proposal C: data-first7,643 words
Curation and evaluation loop first; the live dashboard consumes corpus-derived policy.

Research

Reference copilot dissection10,957 words
How the open-source Jev sales copilot works, line by line, and what transfers.
CurrencyTransfer domain brief12,279 words
Customer journey, product facts, draft question banks and playbooks for onboarding and customer success. Marked DRAFT for review.
Corpus and curation6,056 words
Using the historical call corpus: ingestion, Jev quality rubric, the admin curation loop, storage, evaluation harness.
Speech-to-text options3,877 words
Whisper vs Deepgram nova-3 on Workers AI, diarisation, costs, and the live verification on a real recording.
Prior art7,592 words
Other Jev call copilots, commercial coaching tools, and the ten most transferable ideas.

Reference

Jev builder's guide7,794 words
Request and response contract, semantics, Cloudflare integration, patterns, limits and gotchas. Verified against the vendor docs.

Verified so far

Jev on CloudflareVerified on the CurrencyTransfer account: about 400 ms per request over REST, Worker binding live at jev-lab.currencytransfer.workers.dev.

BillingPrepaid AI Gateway credits, $0.042 per million input tokens, output free. Without credits every call returns HTTP 402.

CorpusAircall coaching database: 1,895 usable onboarding calls (229 hours) and 1,226 customer-service calls (127 hours), diarised and timestamped, already scored on six rubric dimensions.

Speech-to-textDeepgram nova-3 on Workers AI diarises a real CT recording with 94% speaker agreement at about $0.05 per 10-minute call. Historical calls need no STT.

Decisions taken

Milestone 1: upload or pick a recorded call and replay it against a live dashboard: signals, stage, must-say checklist, objection handling, next best move.
Both scenarios from the start: onboarding (the activation call after the sign-up wizard) and customer success.
Headline metric: must-say completeness plus risk flags for onboarding; resolution and call health for customer success. No closing probability.
Suggested lines: pre-approved playbook lines, plus tailored rewrites when Jev verifies them for invented facts, promises and rate predictions.
Curation: Jev scores every historical call; the admin marks Jev-tagged moments as model, acceptable or avoid.