Self-learning Few-shot Text-to-SQL
A retrieval-driven few-shot engine that harvests successful queries as exemplars and injects them dynamically — lifting text-to-SQL accuracy from 60% to 95% with no retraining.
role: Designer & developer
Idea
Instead of fine-tuning, let the system learn from its own successes: every query that produces a validated result becomes a new in-context exemplar.
Result
Accuracy rose from 60% to 95% — with no model retraining or fine-tuning cost — and keeps improving as usage grows.
Tool-Selection LLM — LoRA Fine-tune on xLAM
Teaching a small LLM to pick the right tool/API for an agent — and say how confident it is. LoRA-tuned Qwen3.5-2B with Unsloth on xLAM function-calling data: 50.6% → 100% accuracy, ECE 0.220 → 0.002.
DocuMind AI — Grounded Multimodal Document Q&A at Scale
System design for a 1M-user document Q&A product: async multimodal ingestion, hybrid retrieval, a deterministic confidence gate that refuses instead of hallucinating, and hot/warm/cold session tiering.
Agentic AI Chatbot Platform
Production multi-agent assistant for cybersecurity & case management — a router agent dispatching to five specialist agents across hybrid RAG, text-to-SQL and action execution, on self-hosted vLLM.