# Tony Kipkemboi

> AI engineer building practical agent systems, fine-tuning experiments, and developer tools, then sharing the lessons through field notes, courses, and talks. US Army veteran. Speaker at PyCon US, ODSC, Harvard. Former CrewAI, Snowflake, Bloomberg.

## Links

- [Website](https://tonykipkemboi.com)
- [Blog](https://tonykipkemboi.com/blog): Field notes on AI agents, fine-tuning, and developer tools
- [Learn](https://tonykipkemboi.com/learn): Distilled guides from what I am learning and building
- [Projects](https://tonykipkemboi.com/projects): Open source AI projects
- [Press](https://tonykipkemboi.com/press): Talks, courses, podcasts, and features
- [GitHub](https://github.com/tonykipkemboi)
- [YouTube](https://www.youtube.com/@tonykipkemboi)
- [LinkedIn](https://linkedin.com/in/tonykipkemboi)
- [X/Twitter](https://x.com/tonykipkemboi)

## Full Content

- [Complete AI-readable content](https://tonykipkemboi.com/llms-full.txt): Full machine-readable version with all blog posts, projects, and details

## Learning Paths

- [LoRA 101](https://tonykipkemboi.com/learn/lora-101): A plain-English guide to how Low-Rank Adaptation makes fine-tuning smaller, cheaper, and easier to swap.

## Latest Blog Posts

- [In the Trenches #2: The Shrinking Middle](https://tonykipkemboi.com/blog/in-the-trenches-shrinking-middle): A calm note on the middle layer between end users and builders getting smaller, more technical, and closer to production.
- [In the Trenches #1: The Context Extraction Problem](https://tonykipkemboi.com/blog/in-the-trenches-context-extraction-problem): Useful enterprise AI starts with pulling enough real context out of the workflow before you build.
- [Fine-tuning Whisper on Kalenjin: a $25 LoRA experiment](https://tonykipkemboi.com/blog/fine-tuning-whisper-kalenjin): Whisper does not speak my first language. So I trained a LoRA adapter on it for a weekend, on a $25 GPU budget. Here is everything that broke, every number that surprised me, and the lesson I came away with.
- [Text Measurement Without the DOM](https://tonykipkemboi.com/blog/pretext-text-measurement): A tiny JS library that calculates text height at any width without touching the DOM. Built by the creator of react-motion.
- [AI Agent Skills and Plugins Explained (2026)](https://tonykipkemboi.com/blog/agent-skills-and-plugins-explained): Skills are reusable instructions for AI tools. Plugins bundle them for distribution. Almost every major AI tool uses the same format.

## Top Projects

- [Ollama PDF RAG](https://github.com/tonykipkemboi/ollama_pdf_rag) (527 stars): A locally-hosted RAG (Retrieval-Augmented Generation) system that allows users to chat with their PD...
- [CrewAI Gmail Automation](https://github.com/tonykipkemboi/crewai-gmail-automation) (191 stars): Automate Gmail inbox management using CrewAI agents. Intelligently categorizes, responds to, and org...
- [Resume Optimization Crew](https://github.com/tonykipkemboi/resume-optimization-crew) (151 stars): AI-powered resume optimization system using CrewAI. Analyzes and enhances resumes to match job descr...
- [Trip Planner Agent](https://github.com/tonykipkemboi/trip_planner_agent) (143 stars): CrewAI agents that can plan your vacation. Uses multi-agent collaboration to create detailed itinera...
- [Streamlit Replicate Image App](https://github.com/tonykipkemboi/streamlit-replicate-img-app) (103 stars): Image generation application built with Streamlit and Replicate API. Generate AI images using variou...

## Expertise

AI Agents, CrewAI, LangChain, RAG, Python, Streamlit, Local LLMs (Ollama, Groq), Fine-tuning, Developer Education