shivam.
Open to roles
View work Résumé
Graduate AI Product Lead · University of Rochester · Mar → Dec 2025

Ain Center

VentureForgeAI simulationEntrepreneurship
StrategyDecision practice
SyntheticExploration exercises
ProductScope through rollout
Inside the workflow

Practice the reasoning behind a decision.

  1. Explore

    Begin with a business scenario or venture idea.

  2. Choose

    Make assumptions and tradeoffs explicit.

  3. Reflect

    Use AI feedback to question the reasoning.

  4. Validate

    Turn an exercise into questions for real discovery.

The product judgment

Synthetic personas and simulated interviews can sharpen a hypothesis. Evidence of demand must come from real people.

Illustrated from this case study. Examples are synthetic and contain no customer data.

The short version

AI tools for practicing strategy and exploring venture ideas. VentureForge simulations, structured feedback and self-serve exploration for students and founders.

The full story

Founders and students need ways to practice decisions before those decisions become expensive. At the Ain Center for Entrepreneurship, I worked on AI tools that make that practice more accessible.

VentureForge

VentureForge uses business scenarios to help users work through strategy choices. The goal is to make the reasoning visible: which assumptions support a decision, what tradeoffs it creates and what the user might have overlooked.

AI feedback gives the user another perspective to examine. It should invite reflection and challenge an assumption, rather than turn a generated score into an authority.

Venture exploration

I also worked on tools for exploring ideas through structured scoring, generated personas and simulated interviews. These can help a founder identify questions worth investigating.

The distinction matters: a simulated customer is synthetic. Its response does not establish demand or replace an interview with a real person. The useful output is a better hypothesis and a plan to test it.

My contribution

As Graduate AI Product Lead, I worked across scope, design, build, onboarding and rollout while completing my master’s. I connected the educational use case to an experience students and founders could navigate themselves.

What I learned

A self-serve tool needs more than a working model. Users need to understand what to enter, how to interpret the result and where its limits are. Those explanations are part of the product.

Next case studyUrban OS