Work
Case studies
Products I’ve led at Peacock and Vimeo, and AI systems I’ve built and run. Each page says what was built, what was verified, and what wasn’t.
All case studies
- 01
Janet
2 modelsone plans, one executesMy personal AI operating system: agents with clear jobs, an independent reviewer, and a hard line on anything that leaves my Mac.
Claude CodeCodexSkills - 02
Peacock HDR remote config
3 days → 20 minHDR issue resolutionDefined the video-SDK strategy for a remote-configuration tool.
Video SDKRemote configHDR - 03
Voice Journal
~15 minmemo to entry, first testA voice memo becomes a clean entry, and the tasks and events I mentioned show up in Slack for a one-tap yes.
Parakeet (on-device)SlackTickTick - 04
Peacock VOD startup
5.12s → 3.53sVOD startup timeOwned the roadmap for server-side ad insertion data prefetching for binge watching.
SSAIPrefetchPlayback - 05
Scheduler
11 of 13review findings fixed with testsbook.robgentul.com: a booking page that never sees my calendar. Live in test mode.
Cloudflare WorkersD1Slack Socket Mode - 06
Vimeo OTT APAC launch
$726Kvs. a $600K targetLaunched Roku and Samsung TV apps for APAC with Japanese support.
RokuSamsung TVLocalization
Open source
github.com/rgentul → (opens in a new tab)Personal HQ
A daily status page skill: one refresh recipe, one snapshot, made-up demo data.
Coming to GitHubrgentul/product-team-skillProduct Team Skill
Spec-driven product briefs for PM teams in Cursor. A clean-room rebuild.
Coming to GitHubrgentul/schedulerScheduler
Privacy-first booking: a Cloudflare Worker plus a Mac that does the real work.
Coming to GitHubLet’s talk.
Hiring for an AI product role? Grab 30 minutes.