Prism

Prism: An AI Native Decision Workspace

Comment

Inspiration:

For the past few months, I kept asking myself one question: Should I prepare for GATE or focus entirely on placements? I realized I wasn't looking for someone to make the decision for me. I wanted help understanding why the decision felt difficult in the first place. Existing AI assistants gave good answers, but they often skipped the thinking process. That inspired me to build something that helps people reason through decisions instead of simply recommending one.

What it does:

Prism is an AI native Decision Workspace that helps users understand complex decisions before making them. Instead of providing a final answer, it identifies goals, constraints, assumptions, risks, opportunities, and unknowns. It then explores the decision through dynamic reasoning lenses, generates alternative scenarios, stress tests important assumptions, and ends with thoughtful reflections so users can make informed decisions themselves.

How we built it:

I built Prism using Next.js, React, Tailwind CSS, and the OpenAI Responses API with GPT 5.6. Throughout development, I worked closely with Codex to accelerate implementation, refactor components, improve the architecture, and polish the user experience. The application was deployed on Vercel, with a Demo Mode that allows judges to explore the complete workflow even without API access.

Challenges we ran into:

The biggest challenge was designing an AI experience that encourages thinking instead of simply giving advice. I also had to build a reliable Demo Mode because I ran into API quota limitations during development. Creating a structured reasoning workflow while keeping the interface simple and intuitive required many iterations.

Accomplishments that we're proud of:

I am proud that Prism became much more than another chatbot. It transforms a simple prompt into a structured reasoning workspace and focuses on improving human judgment rather than replacing it. I am also proud of building and deploying a polished, production ready application within the hackathon timeframe.

What we learned:

This project taught me that the hardest part of building with AI is not generating answers, but designing meaningful interactions. I also gained hands on experience with Next.js, deployment on Vercel, prompt design, and collaborating with Codex as an engineering partner throughout the development process.

What's next for Prism:

I want Prism to support richer research, collaborative decision making, and personalized reasoning over time. Future versions could include live web research, saved decision histories, document analysis, and integrations with productivity tools. My long term vision is to make Prism a trusted thinking companion for students, professionals, founders, and anyone facing important decisions.

Built With

Share this project:

Updates