Saurabh Kishore

prof_pic.jpg

Machine Learning Engineer building production-grade LLM and retrieval systems for large-scale intelligence and analytics applications. Experienced across the full ML lifecycle including data pipelines, model fine-tuning, inference infrastructure, and evaluation. Passionate about building scalable AI systems that bridge research innovation with production deployment.

Currently working at JSI in Ottawa, where I built IRIS — a generative AI assistant at the core of JSI’s 4Sight investigation platform. IRIS enables law enforcement and intelligence analysts to perform conversational search across large operational datasets, automatically surfacing context-rich insights about entities, locations, and topics relevant to active investigations. I’m also pursuing my M.Sc. in Computer Science (Machine Learning) at Georgia Institute of Technology.

I hold a double undergraduate degree in Computer Science and Cognitive Science (AI Specialization) from Carleton University.

When I’m not building AI systems, you can find me spending time outdoors, playing basketball/volleyball, and I have a passion for photography and music.

Here are some projects I’ve worked on:

  • IRIS (JSI 4Sight) — Production GenAI assistant enabling conversational search over large-scale intelligence datasets for law enforcement agencies (Agno, LangGraph, LanceDB, vLLM) [+]
  • ClarityAI — AI-powered strategic idea validation tool using a 6-agent pipeline (Agno, LangChain, LanceDB, React) [+]
  • PolyMind — Multi-agent RAG system exploring collective reasoning and bias in AI decision-making (LangGraph, LanceDB, RAGAS, DeepEval) [+]
  • Ai.sle6 — Fully local RAG system for natural language retail product discovery (Llama 3.2, Ollama, LanceDB) [+]
  • DevHelper — Agentic toolkit for evaluating and comparing developer tools via natural language (LangChain, Firecrawl, GPT-4, MCP) [+]
  • Semantic Search — Semantic search engine using OpenAI embeddings, tokenization, and n-gram analysis [+]