DHYAN PATEL

About

AI/LLM engineer who has shipped production RAG chatbots using LangChain/OpenAI with FAISS and FastAPI, focusing on real-world constraints like context windows, concurrency, and latency (reported ~40% latency reduction and <2s average response). Experienced orchestrating AI pipelines with Celery and fault-tolerant long-running workflows with Temporal, and has applied NLP model tradeoff testing (Word2Vec vs BERT) to drive measurable accuracy gains.

Experience

Education

Key Strengths

• Built and deployed production RAG chatbot (LangChain + OpenAI + FAISS + FastAPI) for internal document QA
• Performance optimization under context-window and concurrency constraints (chunking, caching, batching, async pipelines)
• Reduced chatbot latency ~40% while improving response accuracy
• Orchestrated multi-step AI pipelines with Celery (parallel tasks, retries) for RAG query workflows
• Designed fault-tolerant long-running workflows with Temporal (state management, retries, checkpoint resume) reducing compute cost
• Structured AI agent/workflow evaluation with measurable metrics (latency, accuracy, consistency) plus load testing and monitoring
• Improved RAG system latency to under 2 seconds via modular design and testing
• Model selection tradeoff analysis (Word2Vec vs BERT); improved sentiment accuracy by 18% with BERT and recovered latency via batching/caching (25% faster)
• Cross-functional collaboration translating non-technical stakeholder needs (teachers/product leads) into AI chatbot workflows