
Sign up to save your podcasts
Or


Meta is one of the biggest dinosaurs in the park, yet in a matter of months it rebuilt PM top to bottom, and almost none of it was bought. My guest is Jagjit Chawla, a VP of Product at Meta who’s running a growing share of the Facebook app: Feed, Reels, and more recently Search. I worked alongside Jagjit for nearly ten years across Google, Credit Karma, and Meta, and almost nothing is as I remember it. These systems were built by the teams themselves, sometimes in a single evening, on tools you already have. Jagjit walks me through how a product org of a few thousand people learned to move like a startup again.
Key topics:
• Inside Meta’s “no-meeting week” that set the org loose on the tools
• The one job AI can't take: influencing other humans
• Why the PMs winning on Jagjit’s team aren't engineers
• What to do when you’re the bottleneck
• How to become an "AI captain"
• Triaging tens of thousands of bug reports with agents that validate and pre-write the fix
• Why letting AI write code made site incidents spike, and how Meta clamped down
• Building a morning brief that flags the decisions only you can make
• Why the management "compression algorithm" is dead and what replaced it
Timestamps:
00:00 Introduction
05:47 Why the biggest change is pace
07:29 From detailed PRDs to a paragraph, a prototype, and an eval set
09:32 Why the management "compression algorithm" is dead
10:37 Visibility is high, but synthesis is scarce
14:45 "Finding a scissor" when you don't have fancy tools
16:10 A morning brief that flags the decisions only you can make
22:37 When the agent quietly drops your source links
25:53 The one job AI can't take: influencing other humans
30:18 Getting an analyst's answer in five minutes, not 24 hours
33:33 Three ways AI changed the products, not just the process
37:25 Why letting AI write code made site incidents spike
40:39 The no-meeting week and the birth of "AI captains"
44:55 The world's largest Jenga game: why institutional knowledge wins
48:41 Why economics and physics PhDs are "out-PMing" engineers
52:14 Ideas and agency before tools
56:19 "AI lowers the floor and raises the ceiling"
58:40 Closing advice: once you taste it, you can't go back
Brought to you by:
• Glean—Work AI that works
• Framer—Design and build your website effortlessly with Framer Agents
Referenced:
• Claude Code
• Credit Karma
• Ferrari
• Gmail
• Google Chat
• Google Drive
• Indian Premier League (IPL)
• Meta
• Porsche
• Prince of Persia
• TikTok
• Workplace from Meta
• Zoom
Where to find Nikhyl:
• Twitter/X
Where to find Jagjit:
Join The Skip:
• Skip Coach
• Skip Community
Find The Skip:
• Website
• Substack
• YouTube
• Spotify
• Apple Podcasts
Don't forget to subscribe to The Skip to hear me coach you through timely career lessons. Access exclusive sessions from 100+ top product leaders at skip.coach. If you’re interested in joining me on a future call, send me a note on LinkedIn, Threads, or Twitter. You can also email me at [email protected]
By Nikhyl Singhal5
7171 ratings
Meta is one of the biggest dinosaurs in the park, yet in a matter of months it rebuilt PM top to bottom, and almost none of it was bought. My guest is Jagjit Chawla, a VP of Product at Meta who’s running a growing share of the Facebook app: Feed, Reels, and more recently Search. I worked alongside Jagjit for nearly ten years across Google, Credit Karma, and Meta, and almost nothing is as I remember it. These systems were built by the teams themselves, sometimes in a single evening, on tools you already have. Jagjit walks me through how a product org of a few thousand people learned to move like a startup again.
Key topics:
• Inside Meta’s “no-meeting week” that set the org loose on the tools
• The one job AI can't take: influencing other humans
• Why the PMs winning on Jagjit’s team aren't engineers
• What to do when you’re the bottleneck
• How to become an "AI captain"
• Triaging tens of thousands of bug reports with agents that validate and pre-write the fix
• Why letting AI write code made site incidents spike, and how Meta clamped down
• Building a morning brief that flags the decisions only you can make
• Why the management "compression algorithm" is dead and what replaced it
Timestamps:
00:00 Introduction
05:47 Why the biggest change is pace
07:29 From detailed PRDs to a paragraph, a prototype, and an eval set
09:32 Why the management "compression algorithm" is dead
10:37 Visibility is high, but synthesis is scarce
14:45 "Finding a scissor" when you don't have fancy tools
16:10 A morning brief that flags the decisions only you can make
22:37 When the agent quietly drops your source links
25:53 The one job AI can't take: influencing other humans
30:18 Getting an analyst's answer in five minutes, not 24 hours
33:33 Three ways AI changed the products, not just the process
37:25 Why letting AI write code made site incidents spike
40:39 The no-meeting week and the birth of "AI captains"
44:55 The world's largest Jenga game: why institutional knowledge wins
48:41 Why economics and physics PhDs are "out-PMing" engineers
52:14 Ideas and agency before tools
56:19 "AI lowers the floor and raises the ceiling"
58:40 Closing advice: once you taste it, you can't go back
Brought to you by:
• Glean—Work AI that works
• Framer—Design and build your website effortlessly with Framer Agents
Referenced:
• Claude Code
• Credit Karma
• Ferrari
• Gmail
• Google Chat
• Google Drive
• Indian Premier League (IPL)
• Meta
• Porsche
• Prince of Persia
• TikTok
• Workplace from Meta
• Zoom
Where to find Nikhyl:
• Twitter/X
Where to find Jagjit:
Join The Skip:
• Skip Coach
• Skip Community
Find The Skip:
• Website
• Substack
• YouTube
• Spotify
• Apple Podcasts
Don't forget to subscribe to The Skip to hear me coach you through timely career lessons. Access exclusive sessions from 100+ top product leaders at skip.coach. If you’re interested in joining me on a future call, send me a note on LinkedIn, Threads, or Twitter. You can also email me at [email protected]

537 Listeners

2,698 Listeners

1,093 Listeners

2,354 Listeners

3,976 Listeners

228 Listeners

837 Listeners

576 Listeners

5,530 Listeners

143 Listeners

682 Listeners

491 Listeners

34 Listeners

95 Listeners

50 Listeners