Deploy FreeLingo and finish a lesson with saved voice practice
At a glance
FreeLingo's README gives a self-hosting path from provider configuration to a lesson-specific tutor conversation. This workflow is listed under Self-Hosting, Education, Voice.
- Agents / tools
- FreeLingo, Lingu, Docker Compose, OpenAI Speech APIs
SourceBuilding a tool for these workflows? See weekly rates.
What they did
1) Prepare Git, Docker, Docker Compose, and an LLM provider. Clone artcc/freelingo and enter the checkout. Copy .env.example to .env, configure database and Redis passwords plus SECRET_KEY, and set LLM_PROVIDER=openai with OPENAI_API_KEY. For speech without an NVIDIA GPU, set TTS_PROVIDER=openai and STT_PROVIDER=openai using that same key. 2) Run docker compose up -d and open localhost:3000. Register, select your study language, complete placement, and choose a study intensity to generate the roadmap. 3) Complete an unlocked lesson, confirm the voice-practice topic preview, and speak with Lingu about that lesson. 4) Reopen chat history to review the conversation under its lesson topic and inspect progress on the dashboard.
Worth knowing
The first registered account becomes an administrator by default. Remote microphone use requires HTTPS, and production voice sessions need a reverse proxy forwarding /ws/* to the backend. Cloud speech requires an OpenAI API key.
