Skip to content
Sonic
AI
Sonic
AI
Home
Discover
Library
Ask Sonic
Projects
Use with Claude or ChatGPT
Show me around
Request source or feature
The Correct Way to Build and Manage AI Agents in 47 Minutes | Jared Zoneraich, Sonic AI
Home
/
Peter Yang
/
The Correct Way to Build and Manage AI Agents in 47 Minutes | Jared Zoneraich
Peter Yang
Notify me
•
Jul 12, 2026
•
47:12
Interview
The Correct Way to Build and Manage AI Agents in 47 Minutes | Jared Zoneraich
Jared Zoneraich
(Builder-in-Residence, Cognition, guest)
Get the full transcript next time Peter Yang releases an episode
Summary, key quotes, top claims, and the searchable transcript - emailed automatically. No card needed.
Sign up
Executive Summary
Cognition's AI agent, Devin, is architected as a cloud-native, asynchronous system, allowing agents to run for extended periods, close their laptops, and even spawn other 'child' agents to work in parallel.
The strategic vision is not to replace software engineers but to multiply their effectiveness by automating 'brownfield' tasks like code migrations, increasing test coverage, and other developer grunt work, freeing humans for high-level decisions.
A key trend is the increasing automation of agent deployment, with more Devin instances now being launched programmatically or by other agents in response to events (e.g., Datadog alerts) than by human users.
Defensibility in the AI agent space is shifting away from complex prompt engineering towards 'tool engineering' (providing the right tools to a capable model) and a robust go-to-market strategy, such as using Forward Deployed Engineers to embed with customers.
Continue your research
Keep pulling the thread on Jared Zoneraich.
The Cloud-Native Agent Paradigm
Augmenting, Not Replacing, Engineers
12
quotes
Transcript
Key Arguments
Analysis
Quotes & Entities
12
Related
Loading transcript...
Processed Jul 12, 2026
Daily intelligence brief →
yt-dlp + mlx-whisper + Gemini