Habit Builder Agent
by Vancesca Dinh
Vancesca Dinh's agent helps users identify goals and understand the 'why' behind them, grounded in data from the Huberman Lab podcast and medical publications. Pipeline includes RSS feeds, Faster Whisper transcription, Qdrant embeddings, query rewriting for better search, Logfire, Pydantic, and guardrails to keep the agent on-task.
Project from the first cohort of the AI Engineering Buildcamp, by Vancesca Dinh.
The Habit Builder agent is grounded in data from the Huberman Lab podcast and medical publications. The data pipeline includes: downloading RSS feeds, transcribing audio with Faster Whisper, and storing embeddings in a Qdrant vector database. For the agent, she implemented a tool that rewrites user queries in three different ways to improve search results. She used Logfire for logging and Pydantic for structure. A key part of her presentation was the need for guardrails: she found the agent would obediently "draw a cute pig" or translate text into Romanian if asked, so she added checks to keep it focused on the intended use.
Tech stack: RSS, Faster Whisper, Qdrant, query rewriting, Logfire, Pydantic, guardrails.
Related content
Cybersecurity Disclosure Agent
An agent that tracks cybersecurity incidents reported to the SEC, built by Scott DeGeest. Helps supply chain professionals quickly find out if a company has disclosed a data breac…
December 15, 2025
Personalization: Intelligent Email Agent
Asia Amodeo's intelligent email agent helps manage an inbox. It integrates with the Gmail API, indexes emails in Elasticsearch, and offers a Streamlit chat interface to find speci…
December 15, 2025
Building an Agent with Guardrails
AI agents are useful because they can call tools and act on user requests. That same flexibility makes them easy to steer off topic. In this workshop, we build a Data Engineering …
May 26, 2026