Parth Sareen
Ollama
Sessions (2)
Additional Speaker Details
Using publicly available information we constructed an analysis to help you get a feel for this speaker before deciding to attend their session.
Sareen builds the structured-output, function-calling, and agent-harness internals of Ollama, and writes unusually deep, implementation-level essays (sampling/grammar-constrained decoding, agent reliability) grounded in real production debugging — worth attending if you want concrete technique over a high-level product pitch.
Speaking style
Deeply technicalBattle-testedDemo-driven
Recent talks (3)
GitHub
@ParthSareen- ParthSareen/obsidian-rag★ 131Local-first RAG (Langchain) over your Obsidian markdown notes
- ParthSareen/simple-rag★ 65Quickly set up local Retrieval Augmented Generation over PDF and Markdown documents
- ParthSareen/ducky★ 26Natural language to bash commands; run, understand, and copy LLM-generated commands
- ParthSareen/CalendarGPT★ 9Manage your Google Calendar through an AI agent over text
- ParthSareen/watchy★ 4Mini agent to run and chat with background tasks using Ollama
Recent writing (5)
- Creating the Ollama Agent Harness · blog · Jul 2026
- The philosophy of Ollama Launch · blog · Jun 2026
- Yet another generalist · blog · Mar 2026
- Building reliable AI agents · blog · Dec 2025
- Sampling and structured outputs in LLMs · blog · Oct 2025
Podcasts & interviews (2)
- Ollama, Local AI, and Open Agents with Parth Sareen · ODSC AI X Podcast · May 2026
- DevReal: Parth Sareen Interview · DevReal (FunctionalTV / AI by the Bay) · Nov 2025