Undermind

What is Undermind?

Undermind is an AI-powered search tool built to help researchers pinpoint the academic papers they need. Unlike typical search engines that return a list of results, Undermind mimics how a careful researcher would systematically discover information. It examines results in stages, using advanced language models to make decisions along the way, reading through hundreds of papers and adjusting its search as it goes. It then explains why each paper it finds is relevant to your topic.

Who created Undermind?

Undermind was developed by Josh Ramette (CEO) and Tom Hartke (CTO), who hold PhDs in quantum physics from MIT. Their company specializes in AI-powered search algorithms designed to mirror a researcher’s meticulous discovery process, with a focus on accurate, comprehensive results for complex research needs.

How to use Undermind

1. Describe your research query: Explain what you’re looking for as if talking to a colleague in your field.
2. Let Undermind adapt: The tool reads through hundreds of papers, adjusting its search strategy to find relevant results.
3. Be patient: Each search typically takes about 2-3 minutes, allowing the AI agent to identify the best papers.
4. Check the database: Undermind draws from the Semantic Scholar database, so a very specific paper may not be available. If a paper doesn’t appear right away, the search may still be in progress, and you can supplement with other resources to ensure complete coverage.

Key takeaways

Undermind is designed for deep, complex academic searches where accuracy and thoroughness matter more than speed. Its staged, adaptive approach and explanations of relevance make it a useful partner for researchers, though its reliance on the Semantic Scholar database is worth keeping in mind for highly specific queries.