How engineers can manage teams of agents and subagents | Git Sh*t Done LIVE 👾 Episode 3
📅 Thursday, April 23rd
🕕 11:00 AM PST | 2:00 PM EST
📍 Live on Streamyard
55% of engineers now run AI agents regularly. Few know them know how to manage more than one at a time and that gap is growing fast.
Running a single agent is hard enough. Running a team of specialized subagents, each with their own context windows, their own failure modes, and the ability to break each other's work downstream, is a completely different skill.
In this episode of Get Sh*t Done LIVE, we're breaking down how engineers are actually managing agent teams in production. How to delegate tasks to subagents without losing control of the output. How to set up orchestrator-worker patterns that hold up at scale and how to keep context from getting lost between agent handoffs.
In this week's episode sponsored by Optimal AI, we're chatting the latest and greatest in agents, an Engineering Leader from HubSpot.
Your hosts: Iba Masood & Syed Ahmed (aka The Gitfather), serial founders, YC alums, and the team behind Optimal AI. Between them, they've spent 15+ years building in AI, and have interviewed over 300 engineering leaders.
The event is ONLINE — you can join from the Streamyard link.
☕️ The Coffee (It’s on Optimal!)
We're picking a few lucky attendees to receive a curated, specialty coffee kit delivered straight to them, alongside a free, unlimited trial of Optibot!
How to get it: Simply RSVP with your shipping info to be entered into the draw.
The Roaster: Kitty Town Coffee.
The Blend: A specialty grade mix from Brazil and Costa Rica. FYI - Every bag shipped feeds a homeless kitty. 🐈
About Optimal AI
Optimal AI is the team behind Optibot; AI engineers that review code, proactively fix CI failures and catch bugs. They identify 2x more security vulnerabilities and tech debt, with full architecture and codebase context vs code review tools. Optibots can pair program with your team and find dead code, compliance issues and vulnerabilities proactively, while reviewing engineering efficiency metrics to understand overall productivity.
Used by teams like MongoDB, MPulse, Prometric and more to save ~4 hours per engineer per week on reviews, security, and compliance work, and help visualize cycle times and productivity.