# Amass

> Amass is the research engine for life science. It unifies 500M+ scientific documents — papers, trials, patents, regulatory filings — and answers questions with cited, source-backed responses.

Amass is built for biotech, pharma, and research teams. Every claim is traceable to a primary source. Available as a web app, an MCP server for use inside AI tools, and a developer API.

Every page below links its markdown variant directly (the `.md` suffix); the same HTML page lives at the URL without the suffix, and full concatenated content is in [llms-full.txt](https://amass.tech/llms-full.txt).

## Product

- [Product overview](https://amass.tech/product.md): The three ways to work with Amass — App, MCP, and API. One research engine, three surfaces.
- [Amass App](https://amass.tech/app.md): Ask a research question in plain language. Amass works through it across the eight data layers, shows every search it ran, and cites each claim to a source you can open. Longer runs produce downloadable, cited reports.
- [Amass Core](https://amass.tech/core.md): The specialised data layers that power retrieval — biomedical literature, clinical trials, drugs, genes, patents (preview), and FDA/EMA regulatory. All live on the public API today.
- [Amass MCP](https://amass.tech/mcp.md): Use Amass inside the AI tools you already use (Claude, Cursor, ChatGPT, etc.) via the Model Context Protocol.
- [Amass API](https://amass.tech/life-science-api.md): Developer access to the same retrieval engine that powers the App and MCP.

## Use cases & content

- [Use cases](https://amass.tech/use-cases.md): Concrete workflows — competitive landscaping, target validation, regulatory monitoring, and more.
- [Resources](https://amass.tech/resources.md): Guides, articles, and documentation.
- [How to do scientific work in Claude](https://amass.tech/how-to-do-scientific-work-in-claude.md): Using Amass inside Claude — biomedical literature, clinical trials, and regulatory data, with every answer cited to source.
- [Skills & Prompts](https://amass.tech/skills.md): Ready-made skills and prompts that turn Amass into finished analyst work inside Claude, Codex, and other assistants, each with a validated answer and downloadable outputs.
- [Build an AI scientist](https://amass.tech/build-an-ai-scientist.md): How to build an agentic, source-grounded research workflow instead of one-off prompts.
- [DTU Skylab hackathon](https://amass.tech/dtu-hackathon.md): Amass at the AI in Life Sciences hackathon, DTU Skylab, 16 September 2026. Covers API credits, five project ideas, and a copy-paste recipe for each.

## Company

- [About Amass](https://amass.tech/about.md): Team, mission, and how the company is built.
- [Security](https://amass.tech/security.md): Infrastructure, data handling, and enterprise terms.
- [Pricing](https://amass.tech/pricing.md): Plans and limits.
- [Services](https://amass.tech/services.md): Training, implementation and forward-deployed engineering — including custom *Core builds.
- [Careers](https://amass.tech/careers.md): Working at Amass — Copenhagen-based.
- [Contact](https://amass.tech/contact.md): Get in touch.
- [Book a demo](https://amass.tech/demo.md): Schedule a walkthrough with the team.

## Legal

- [Privacy Policy](https://amass.tech/legal/privacy-policy)
- [Terms of Service](https://amass.tech/legal/terms-of-service)
- [API Terms of Service](https://amass.tech/legal/api-terms-of-service)

## Machine-readable

- [llms-full.txt](https://amass.tech/llms-full.txt): Every page's full markdown content in one file.
- [Sitemap](https://amass.tech/sitemap.xml)
- [Schemamap](https://amass.tech/schemamap.xml): Per-page JSON-LD endpoints (each page also serves its graph at the `.jsonld` suffix).
- [MCP server manifest](https://amass.tech/.well-known/mcp.json): The Amass MCP server — streamable HTTP at https://mcp.amass.tech/mcp.
- [Agent catalog](https://amass.tech/.well-known/ard.json): ARD entry for the MCP server.
- [agents.txt](https://amass.tech/agents.txt): Agent capability declaration.