Announcements

What does a 10x researcher workflow look like in practice?

In the recent research paper Tadpole Nahm sum as a Wronskian, available at arXiv, Shane Chern (Postdoctoral Researcher, University of Vienna) and MathVision cofounders Chanh Tran and Tanay Wakhare used MathVision to solve an important open math problem.

The key breakthrough came when MathVision surfaced an overlooked reference to research of Bartlett and Warnaar. Shane Chern, a combinatorialist specializing in partition theory and q-series, was able to instantly realize the importance of this reference.

Shane said: “It was really amazing to see MathVision help us discover the overlooked connection between tadpole Nahm sums and the Bartlett-Warnaar series. I liken it to the modern Baker Street Irregulars.”

Once this connection appeared, using the MathVision workspace we were able to quickly investigate the relevant identities, iterate on the argument, and turn the observation into a key lemma of the paper. This is the workflow MathVision is built for: problem → literature → iteration → final proof → LaTeX → verification.

Our goal is to build the first AI-native workflow automation tool for mathematical research and enable the 10x researcher, helping mathematicians find the right ideas faster and iterate dramatically more efficiently. MathVision is building cutting-edge computational tools in order to transform pure and applied research workflows.

MathVision is currently free. Try it at mathvision.ai.

We’re excited to introduce MathVision, an all-in-one workspace for AI-assisted mathematical and applied research.

MathVision takes currently scattered computational experiments, chat windows, Overleaf writeups, and notes, and unites them into a single research workspace. Our AI-assisted workflows can accelerate parts of the research cycle by up to 10x while leaving you in control. Sign up today at mathvision.ai.

MathVision is built for:

• Graduate students and postdocs who want access to a math-native research assistant.
• Applied researchers who want a better workflow than scattered computational experiments, chat windows, Overleaf writeups, and notes.
• Professors exploring how AI can support serious research without complicated setup.

Our platform focuses on:

• Proof exploration: using LLMs for testing ideas, clarifying assumptions, finding gaps, and quickly exploring proof strategies.
• LaTeX writing: drafting and polishing mathematical text for more publication-ready text.
• End-to-end workflows: tracking intuition, failed approaches, promising ideas, and human feedback in one place.
• Agentic workflows and code integration: quickly run semi-autonomous computational experiments and analyze the results.
• Formal verification readiness: incorporating cutting-edge Lean verifiers for provably correct mathematics.

Our founding team unites deep academic expertise with industry experience:

• Tanay Wakhare (PhD candidate at MIT), CEO
• Rimon Melamed (PhD candidate at GWU), CTO
• Chanh Tran, COO, Head of Product

MathVision is currently being released free of charge. We want serious AI assistance to be accessible to mathematicians and researchers without expensive subscriptions or specialized tooling. We are already used by researchers at MIT, Harvard, Stanford, Princeton, and other leading global institutions.

We have years of AI consulting experience having founded Prompt Inversion, where we’ve helped organizations implement secure and scalable AI systems across 10+ industries. We know how to implement AI systems that work in the real world. MathVision brings that same experience to mathematical research.