Solvita

Inspiration

Learning math alone is difficult when you don’t know what to study or whether your solutions are correct. Private tutors are expensive, inconsistent, and parents rarely know how their children are actually performing. We wanted a smarter, affordable system that guides students logically and gives parents real insight.

What it does

Solvita lets students solve math problems step-by-step and get AI validation on their reasoning, not just the final answer. Parents can schedule sessions, track completion, and receive objective updates on their child’s focus, effort, and progress.

How we built it

We built Solvita using Cursor, Claude Code, Lovable, Perplexity for research, and N8N for automation. The system was assembled iteratively — backend logic, agent workflows, and UI evolved as we tested and refined the idea.

Challenges we ran into

We struggled with consistent LaTeX formatting, deciding which features actually belonged in the UI, and coordinating multiple agents with different tools. Integrating everything into a stable workflow took more time than expected.

Accomplishments we’re proud of

We built a clear and functional backend, a clean MVP UI, a working parent dashboard, automated agents powering the system, and a coherent brand and logo.

What we learned

We got better at building AI agents with Lovable, automating flows with N8N, and designing a simple, understandable MVP. We also discovered new tools that improved how we built and presented the project.

What’s next for Solvita

Next steps are user testing, expanding our exercise library, and polishing the full learning flow. We’ll also extend Solvita beyond math into other subjects, building a suite of AI tutors — each focused on a specific domain, all following the same principle: guided reasoning, actionable feedback, and transparency for parents.

Built With

• claude
• cursor
• lovable
• n8n
• supabase