Building Effective Agent Harnesses

Hosted by Hugo Bowne-Anderson

Thu, Jul 23, 2026

10:00 PM UTC (30 minutes)

Virtual (Zoom)

Free to join

Invite your network

Go deeper with a course

Build AI Agents from First Principles
Hugo Bowne-Anderson
View syllabus

What you'll learn

See what turns an LLM call into an agent

Build the core loop: give the model tools, let it act, return the result, and repeat until the task is done.

Build the harness around the model

Use context, state, guardrails, traces, MCP, hooks, cron jobs, sub-agents, and feedback to shape reliable behavior.

Steal the right ideas from modern agent harnesses

Study Pi, Codex, and Claude Code, then decide which ideas your users need and which belong on a conference slide.

Why this topic matters

Agent frameworks have an impressive talent for making every feature feel mandatory. Memory! Planning! Multi-agent orchestration! You ain't gonna need most of it. We’ll build an agent loop from first principles, examine MCP, hooks, cron jobs, and sub-agents, and learn from Pi, Codex, and Claude Code so you can decide what belongs in your harness and what belongs on a conference slide.

You'll learn from

Hugo Bowne-Anderson

AI & data engineer, consultant, educator of 6+ million students (ex-Yale)

Hugo Bowne-Anderson is an independent data & AI consultant with extensive experience in the tech industry. He has advised and taught teams building AI-powered systems, including engineers from Netflix, Meta, and Amazon. He is the host of Vanishing Gradients, where he explores developments in data science and artificial intelligence, and has written for publications including Harvard Business Review and VentureBeat. Previously, Hugo served as Head of Developer Relations at Outerbounds and held roles at Coiled and DataCamp, where his work in data science education reached over 6 million learners. He has taught at Yale University, Cold Spring Harbor Laboratory, and conferences including SciPy and PyCon, and is a passionate advocate for democratizing AI skills and open-source tools.

Coached teams at

Google
Instagram
OpenAI
Netflix
Yale
See all products from Hugo & Stefan

Sign up to join this lesson

By continuing, you agree to Maven's Terms and Privacy Policy.