Skip to content

v0.1.0 — first release

Latest
Latest

Choose a tag to compare

View all tags
@ngoanpv ngoanpv released this 24 Jun 09:07
v0.1.0
a2f1f9e
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

First tagged release of DeepInterview — an open-source (Apache-2.0), voice-first AI mock-interview platform.

What works end-to-end

The full closed loop runs against real providers: CV + JD → company research → adaptive real-time voice interview → scorecard/report → study coach.

  • English-first, multilingual — language is a per-session setting; STT/TTS/voice route by language.
  • Cascaded STT → LLM → TTS for transcripts + per-component control.
  • Runs offline on mock providers with zero keys, or on real providers via deepinterview init + docker compose up.
  • Prep (LangGraph) and scoring pipelines, LiveKit live interviewer, study coach + knowledge sidecar, avatars, and a multi-agent design.

Highlights since pre-release

  • Verified full-stack, real-provider run of the prep → score pipeline on the dockerized stack.
  • fix(prep): recover the real question plan when the model appends trailing JSON — previously the keystone planner could silently fall back to a generic plan.
  • feat(cli): deepinterview init is now a real guided key wizard (pick a run mode, enter only the keys you need) — the adoption lever for clone-and-run.

Get started

git clone https://github.com/ngoanpv/DeepInterview && cd DeepInterview
pnpm install && pnpm build
pnpm deepinterview init      # guided setup (or run offline with mock providers)
docker compose up            # full stack
Assets 2
Loading