Chatty AI
Multi-model AI streaming platform integrating 10+ models via unified API architecture
Overview
A multi-model AI platform integrating 10+ foundation models (OpenAI, Claude, Gemini) via a unified API architecture. Features real-time Server-Sent Events (SSE) streaming delivering sub-500 ms first-token latency, comparative model output evaluation, and production deployment on Vercel.
Key Highlights & Impact
- Developed a multi-model AI platform integrating 10+ models via unified API architecture.
- Achieved sub-500 ms first-token latency using real-time streaming (SSE).
- Managed production deployment on Vercel, including routing, environment configuration, and performance tuning.
- Engineered interactive prompt comparisons and deterministic guardrails.