AI Portfolio Platform

Static portfolios fail to showcase deep technical expertise interactively. Recruiters need contextual answers about experience.

Built a RAG pipeline with semantic search, prompt engineering, and pluggable LLM providers with real-time SSE streaming.

Next.js frontend → ASP.NET Core API → Semantic Search → Qdrant → Prompt Builder → OpenAI → SSE streaming with citations.

• Balancing retrieval quality with latency budgets
• Supporting multiple LLM providers with a unified interface

• Chunking strategy dramatically affects answer quality
• Observability across the RAG pipeline is essential