High-Stakes AI Product Decisions
Prachie Banthia
Ex-VP Product AssemblyAI & YC founder
Lead AI product decisions without false precision
AI changes how product leaders are judged.
Shipping is faster. Prototypes are cheaper. Demos are easier. What’s harder now is deciding what’s actually worth building and defending that decision when the system itself is nondeterministic.
Most product leads feel this gap:
Customer calls sound positive, but not decisive
ROI models feel fake, but leadership and investors still expect numbers
Feasibility shifts as models and costs change
You’re asked to project confidence without overstating certainty
A demo works once, fails the next time, and no one knows if that’s signal or noise
As a result, teams drift in endless validation, overbuild impressive demos, or make decisions they can’t clearly explain later.
This course exists to fix that.
You’ll learn how to make high-stakes AI bets without false precision. You’ll learn how to turn unclear customer calls, shifting tech capabilities, and incomplete data into a clear go/no-go decision you can defend.
You’ll leave with an exec-ready decision memo for a real AI bet and a repeatable way to make these calls again.
In my experience across big tech and startups, this skill separates teams shipping AI slop from teams leading real AI strategy.
What you’ll learn
Learn how to frame, decide, and defend AI product bets under real uncertainty.
Define the decision boundary when capabilities, costs, and quality are still moving
Make explicit what assumptions might break as models or infra change
Produce a decision statement that holds even as the tech shifts
Identify which customer insights actually move a strategic decision
Discord qualitative inputs that sound positive but don't affect commitment
Translate messy feedback into evidence leadership can act on
Define what evidence should trigger a decision revisit versus what is expected model variance
Avoid reopening decisions due to noise, novelty, or tech change
Keep teams aligned as models, quality, and costs evolve
Identify where pricing uncertainty meaningfully changes the bet
Compare options under different pricing and margin scenarios
Prevent teams from stalling on pricing just because no one understands it
Learn directly from Prachie
Prachie Banthia
YC Founder and former Product Exec at AssemblyAI, Gopuff
Who this course is for
Senior PMs who ship AI features well and are stepping into owning and defending high-stakes AI product bets.
Product leaders in or preparing for executive-level Product roles that are now accountable for defending decisions in exec or board rooms.
Founders leading AI products who are moving from intuition-led bets to defensible, decision-driven strategy.
Course syllabus
6 live sessions • 13 lessons • 2 projects
Week 1
Feb
17
Frame the Decision Like an Investor
Tue 2/175:00 PM—6:30 PM (UTC)Feb
19
Product Strategy When Model Quality Won’t Hold Still
Thu 2/195:00 PM—6:30 PM (UTC)
Frame the Decision Like an Investor
Feb
18
Optional: Office Hour with Prachie 1
Product Strategy When Model Quality Won’t Hold Still
Week 2
Feb
24
Stakeholder Conversations That Change Decisions
Tue 2/245:00 PM—6:30 PM (UTC)Feb
26
Monetization Strategy for Decision-Makers
Thu 2/265:00 PM—6:30 PM (UTC)
Stakeholder Conversations That Change Decisions
Feb
25
Optional: Office Hour with Prachie 2
Monetization Strategy for Decision-Makers
Schedule
Live sessions
3-4 hrs / week
Wed, Feb 18
5:00 PM—6:00 PM (UTC)
Wed, Feb 25
5:00 PM—6:00 PM (UTC)
Tue, Feb 17
5:00 PM—6:30 PM (UTC)
Projects
2-6 hrs / week
Work on a capstone project for your own real product. This is a decision memo you can use at work right after the course ends.
Async content
2-6 hrs / week
You get lifetime access to the recorded content, so that you can learn at your own pace.
$1,500
USD