Undermind
Undermind is an AI co-researcher and literature layer for R&D that finds highly relevant scientific papers, follows citation trails, reads full texts, and helps researchers build understanding over time. Its goal is to make complex, high-stakes scientific discovery much more effective than keyword search.
At a Glance
- Research scientists
- Doctors and clinical researchers
- Biotech and pharmaceutical R&D
- Academic and university research groups
- +4 more
AI Tools by Undermind
(1)
Undermind
AI Scientific Literature Search
Discussions
No discussions yet
Be the first to start a discussion about Undermind
Latest News
Undermind publishes a current deep-search benchmark reporting 73% recall at 10 minutes and materially higher recall than the compared frontier web-search agents.
Undermind promotes its enterprise deep-literature-search offering for R&D teams and their agents.
Undermind announces/identifies its membership in the Genesis Mission Consortium.
Undermind highlights MCP connectivity to external AI agents and products including Claude, ChatGPT, Copilot, Gemini, Codex, Cursor and Perplexity.
Products & Services
An AI research agent/search engine for complex scientific questions. It asks clarifying follow-ups, searches and evaluates hundreds of papers, follows citations, synthesizes reports, supports full-text and figure reading, and provides inline source citations.
A way to connect Undermind's literature-search capabilities to external AI agents and products, including Claude, ChatGPT, Microsoft Copilot, Gemini, Codex, Cursor and Perplexity.
Enterprise offering with increased compute, organizational login, onboarding seminars, admin dashboard, custom terms/SLA/security review and dedicated support.
Market Position
Undermind positions itself as an AI literature layer for R&D and a specialist alternative to Google Scholar, PubMed and general-purpose web-search agents. Its differentiation is adaptive, citation-aware deep search for complex questions, full-text/figure analysis and longitudinal research context; YC claims 10–50x better results than traditional keyword search, while the current benchmark reports 73% recall at 10 minutes versus 29–39% for the compared frontier web-search agents.
Leadership
Founders
Joshua Ramette
CEO and co-founder; MIT PhD in quantum computing architecture design, with research experience spanning gravitational-wave astrophysics (LIGO), particle physics (CERN), atomic physics and quantum information science. YC describes him as having six peer-reviewed publications and three patents pending.
Thomas (Tom) Hartke
CTO and co-founder; studied physics at Princeton and earned a PhD in quantum physics at MIT. His atomic-physics thesis was recognized by the American Physical Society as one of the field's top four theses in 2022; he has published 10 papers, including two in Nature and Science, and has experience building and deploying complex systems.
Executive Team
Joshua Ramette
CEO
MIT quantum-computing PhD and former researcher across LIGO gravitational-wave astrophysics, CERN particle physics, atomic physics and quantum information science.
Thomas (Tom) Hartke
CTO
Princeton physics alumnus and MIT quantum-physics PhD; atomic-physics researcher and builder of complex deployed systems, with publications in Nature and Science.
Board of Directors
Founding Story
The founders are two MIT quantum-physics PhDs who experienced the frustration of spending hours or months searching scientific literature and missing the one paper that could change a research plan. They started Undermind to rebuild search around an AI agent that adaptively explores papers like a careful human researcher, rather than relying on short keyword queries.
Business Model
Revenue Model
Freemium subscriptions for individual researchers, per-person team subscriptions, and custom-priced enterprise agreements with dedicated support and security/SLA terms.
Pricing Tiers
For exploratory research; includes AI chat, deep searches, reports, shared workspaces, agent connections and standard rate limits.
For serious ongoing research; latest AI models, deepest full-text analysis, 10x higher usage limits, unlimited workspaces, files and paper libraries.
Five seats at $15 per person; adds team-member management, priority support and centralized billing.
Custom terms/SLA/security review, increased compute, organizational login, onboarding, admin dashboard and dedicated support.
Target Markets
- Research scientists
- Doctors and clinical researchers
- Biotech and pharmaceutical R&D
- Academic and university research groups
- Machine-learning researchers
- Engineering and hard-tech teams
- Scientific literature reviews
- Drug discovery and biomedical R&D
- Finding obscure or highly specific papers
- Technical research and engineering discovery
- Research-program planning and direction setting
- Keeping teams current with new publications
- GSK
- Novartis
- Mayo Clinic
- Bain Capital