Show HN: Build AI chatbots and structured APIs easily with custom RAG knowledge

Hey HN! I built this because managing multiple AI providers and getting structured outputs was painful.

What it does:

• Chatbots: Upload docs/PDFs, they’re vectorized into a RAG knowledge base. Write instructions, pick any model (GPT-4, Claude, Gemini, etc), get a working chatbot with iframe embed in ~30 seconds.

• Structured APIs: Define JSON schemas for outputs. Create functions that take any input type (images, PDFs, video, audio, text) and return your structured data. Switch AI models anytime without code changes.

• Custom Knowledge: Vector DB for RAG across all your tools. Upload once, use everywhere.

Real examples: Nutrition extraction from food photos, form validation against company policies, batch PDF data extraction, image analysis with structured outputs.

Dev features: Schema versioning, A/B test different models (cost vs quality), unlimited API keys with per-key analytics, OpenAPI specs auto-generated, full observability.

Everything is hosted and production-ready immediately. One subscription gets you access to all flagship AI models.

Happy to answer questions! Feedback very welcome. Live: easyai.passiolife.com