API / Developers

Built to be replaceable, not locked in

Every external capability sits behind an interface, so demo providers can be swapped for production APIs without touching product code.

Provider interfaces

TranscriptionProvider

Speech-to-text with timings, language detection and speaker hints. Swap in Whisper or an equivalent.

ScriptProvider

Transcript in, rewritten script out, plus hooks, titles, descriptions and hashtags.

VoiceProvider

Text-to-speech across voices, languages and styles. Swap in ElevenLabs or an equivalent.

MediaProvider

AI B-roll/image generation and licensed stock search with license metadata.

RenderProvider

FFmpeg-based composition, encoding, caption burn-in and audio mixing.

PaymentProvider

Stripe or Razorpay behind one interface, selected by configuration.

Asynchronous pipeline

No blocking requests
  1. 01

    Upload to private storage

  2. 02

    Create job record

  3. 03

    Queue job

  4. 04

    AI processing (transcribe → analyze → script → voice)

  5. 05

    Video rendering (FFmpeg)

  6. 06

    Store output privately

  7. 07

    Update job status

  8. 08

    Notify user

Public API

A token-authenticated REST API for creating projects, starting jobs and fetching signed export URLs is planned. Secrets are always read server-side from environment variables — never from browser code.