Open to opportunities

Piyush Desai

@piyushdesai

Message

I build production Generative AI, RAG, and anomaly detection systems.

India

Message

What I'm looking for

I'm looking to build practical Generative AI, RAG, model-serving, and machine learning systems where I can deepen my production experience and contribute to real-world automation, monitoring, and intelligent workflows.

At Paryai Solutions, I'm building production AI systems for document Q&A, sensor monitoring, and automated alerting. I developed an offline RAG pipeline using Ollama, hybrid BM25 and vector retrieval, cross-encoder reranking, and Hugging Face embeddings.

I've also built statistical and multivariate anomaly detection for tank-level sensors and blower equipment, using Isolation Forest to identify mechanical faults and inconsistent readings. My online-learning inflow forecasting work uses River, real-time deviation monitoring, and four-hour sliding-window predictions to trigger sustained-anomaly alerts.

I'm a Computer Science Engineering graduate who enjoys applying Python, FastAPI, open-weight LLMs, FAISS, and machine learning to practical production problems. My projects include predictive maintenance for wind turbines and a house-price regression model designed for REST API integration.

Experience

Work history, roles, and key accomplishments

Current

Python Developer Intern

Current

Paryai Solutions

Feb 2026 - Present (6 months)

Built a production offline RAG pipeline for document Q&A using local LLMs (Ollama), hybrid BM25 + vector retrieval, and cross-encoder reranking with Hugging Face embeddings. Developed statistical and multivariate anomaly detection for sensor data, and implemented online-learning inflow forecasting with automated alerting.