Agents in Production 002: AI code you can trust

​Fast code is the easy part. Knowing it works is harder. When agents write more of your code, your review process becomes the bottleneck, and a confident answer can still be wrong.

​We’ll cover:

• ​How to set up agents that catch and fix their own failures

• ​Tests, benchmarks and AI code review that stop a bad merge

• ​How to test an agent on simulated users before real ones reach it

• ​What to automate first, and what never to hand over

• ​Which architecture calls should stay with a human

​Who it’s for: CTOs, engineering and product leaders at healthcare, hospitality and retail companies with 20 to 500 people.

​Bonus: we’ll review the agent setups of 5 attendees and send each a one-page list of what would break first and ways to manage it.

​Speakers
Mykhailo Kushnir, co-founder of DestiLabs, leads development and delivery of the company’s AI agents. He has spoken at LLM Day, EcomExpo Vilnius and NextApp in Berlin.

​About DestiLabs
Delivered 40+ AI projects in the last 3 years for companies such as CareRx, Composite Global Partners, Stitch.ai and others. Clutch top 100 AI agent dev companies. In this free live series, we show what it takes to run AI agents in production.

​Seats are approved by hand, and you’ll hear from us within 24 hours.

​Format: 60 minutes, with live demos and open Q&A. This session is recorded.

​Also in the series: 001: Everyone’s running AI agents. Who’s in charge?, Nov 18 and 003: Keep them cheap and running, Dec 9