Intuify

Inspiration

Physics problems are often hard not because of the math, but because it’s difficult to visualize what’s actually happening. We wanted to close that gap between equations and intuition by turning problems into something you can see, manipulate, and understand interactively.

What it does

Intuify converts physics word problems into interactive simulations. Instead of just solving for an answer, users can visualize the system, adjust variables, and immediately see how the physics responds in real time.

How we built it

We built a pipeline that translates natural language into a structured physics representation, which then drives a real-time simulation engine. This required mapping text to variables, constraints, and equations in a consistent way across different systems.

We also integrated MCP-compatible tooling, allowing AI agents like Claude to generate and interact with simulations directly.

Challenges we ran into

The biggest challenge was reliably parsing messy, real-world input. Users don’t write clean, structured physics problems, so we had to interpret inconsistent phrasing, missing variables, and ambiguous units, then map that into a valid physics model without making incorrect assumptions.

Another major challenge was supporting combinations of physics systems. When multiple systems interact, such as an inclined plane connected to an Atwood machine, we had to coordinate multiple equations, shared variables, and constraints while keeping the behavior physically consistent.

We also had to ensure that the visuals stayed aligned with the underlying physics. As complexity increased, keeping animation, scaling, and labeling accurate required careful iteration.

Accomplishments that we're proud of

We built a system that turns plain English into fully interactive physics simulations, not just static solutions.

We’re especially proud of supporting combined systems, where multiple physics concepts interact in a single simulation, allowing users to explore how different principles work together rather than in isolation.

What we learned

We learned that clarity matters more than complexity. Building reliable AI systems requires strong validation, not just generation. We also saw how powerful interactive learning can be compared to static explanations.

What's next for Intuify

We plan to expand to more physics systems, improve parsing accuracy, and develop Intuify into a full learning platform that integrates directly with AI tutors and educational tools.

Built With

• canvas-api
• claude-desktop
• custom-physics-simulation-engine
• grok-api
• mcp
• natural-language-parsing-pipeline
• next.js
• react
• tailwindcss
• typescript