Subvolts take Build reliable AI agents using MLOps principles and new tools like Kitaru. Jul 9, 2026 - Developers
- Learn MLOps principles and tools like Kitaru to build reliable, scalable AI agent systems for production.
- Business
- Understand how to move AI agents from demos to robust systems, ensuring reliability and scalability for your applications.
This episode offers a practical guide for engineers on deploying AI agents, a key focus for major AI labs.
Building Durable AI Agents
Audio streams from Daniel Whitenack and Chris Benson. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
This episode of Practical AI features Hamza Tahir discussing how MLOps principles can make AI agents reliable for production. The conversation covers workflows, agent harnesses, fleets, and infrastructure for scalable systems. It also explores open source tools, production challenges, and ZenML's Kitaru project for building resilient agent systems.
This episode explains how to make AI agents work dependably in real-world applications, not just in tests. It covers the tools and methods needed to build AI systems that are robust, scalable, and easy to manage, which is important for businesses wanting to use AI effectively.
This episode is for developers and engineers aiming to deploy AI agents beyond simple demos. It bridges the gap between generative AI and robust MLOps practices, a critical area as frontier labs like Google Gemini and OpenAI focus on agentic capabilities. Understanding agent harnesses and fleet management offers insights into building scalable, production-ready AI systems, relevant to anyone working with or planning to deploy advanced AI applications.
- HostDaniel Whitenack
- HostChris Benson
- GuestHamza Tahir
- Guest: Hamza Tahir
- Title: Building Durable AI Agents
- Show: Practical AI
- Length: 47 minutes
- Project Mentioned: Kitaru
Chapters from the publisher ↗ · Transcript ↗ Key moments and transcript search come with a later piece.
What does it take to move AI agents from demos to reliable production systems? In this episode, Hamza Tahir explores how MLOps principles are shaping the future of generative AI, covering workflows, agent harnesses, fleets, and the infrastructure needed to build durable, scalable systems. The conversation dives into open source tools, production challenges, and how ZenML's new project, Kitaru, helps developers build resilient, replayable, and observable agent systems.
Featuring:
Links:
Sponsors:
- Framer: The enterprise-grade website builder that lets your team ship faster. Get 30% off at framer.com/practicalai
Upcoming Events:
- Register for upcoming webinars here!
- Midwest AI Summit 2026
- MLOps
- Hamza Tahir
- Practices for managing machine learning models in production.
- agent harnesses
- Hamza Tahir
- Frameworks or tools that help manage and run AI agents.
- fleets
- Hamza Tahir
- Groups of AI agents working together.
44 minNarrative intelligence and the human advantage
WithReed Frerichs
47 minOpen models and the future of Physical AI with NVIDIA
WithMing-Yu Liu
49 minFrom AGENTS.md to Enterprise Deployment
WithNick Kuhn
45 min
1 h 03AI Pioneer Jürgen Schmidhuber: AI Already Feels Pain, Loves, and Is Self-Aware
WithJürgen Schmidhuber
56 minWhy a Nation Can't Outsource Its Frontier AI - Alistair Pullen (Cosine AI)
WithAlistair Pullen
Questions people ask
- What is the goal of building durable AI agents?
- The goal is to move AI agents from simple demonstrations to reliable production systems that can be scaled and managed effectively.
- What is Kitaru?
- Kitaru is a new project from ZenML designed to help developers build resilient, replayable, and observable agent systems.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Daniel Whitenack and Chris Benson.