The Causaly platform

The AI platform for scientific evidence and reasoning

Causaly combines 570+ million biomedical facts and licensed full text with AI that reasons through the evidence, so R&D teams can investigate a question, evaluate options against their own criteria and take an evidence-linked recommendation to review.

See how it works

How does Causaly work?

Causaly works through a query the way a research team would, from defining exactly what is being asked to delivering an answer that states what the evidence supports, what contradicts it and highlights where there are gaps.

Step 01

Frame the question

Causaly recognizes every gene, protein, disease, drug and compound in a question, including chemical structures, and maps each one to a single entity using public vocabularies, curated ontologies and, where needed, your internal data.

Step 02

Plan the investigation

Agentic Research splits a complex question into sub-questions, sends each to the best source and carries what it learns into the next step, recording a reasoning trail that reviewers can follow.

Step 03

Retrieve the evidence

Each sub-question draws on literature, licensed full text, patents, clinical trials, Bio Graph, Pipeline Graph, specialist data sources and a live web search. The Scientific Information Retrieval System maps them to the same entities and puts the most relevant evidence first.

Step 04

Screen and weigh

Documents that mention the topic but do not address the question are set aside. Extended Reasoning weighs the rest, separating evidence that supports each claim from evidence that contradicts it and flagging gaps or conflicts.

Step 05

Apply your method

Scientific Workflows turn a team's best practice into a governed process that runs the same steps to the same evidence standards every time, with an auditable record of each run. Start from the Workflow Library or build your own.

Step 06

Deliver the work

Findings arrive as a PDF, slide deck, chart or table, with every claim and data point linked to its source. They live in shared Workspaces, with notifications when new evidence changes an answer the team is tracking.

How is Causaly different from other scientific AI tools?

Each product is built for different work. Causaly maps cause and effect across the biomedical literature and applies your team's own methods, evidence standards and review criteria, producing work that scientists can inspect, challenge and use.

Causaly
Evidence and reasoning platform
Claude Science
AI science workbench
BenchSci
Preclinical platform
Does it cover the whole R&D lifecycle, from target biology to medical affairs?
✓
Discovery, clinical development and medical affairs
From target identification and indication expansion to analyzing clinical trial data, comparing endpoints and populations, and reviewing safety signals and competitive landscapes
◐Depends on the user's setup
◐Preclinical through IND
Does it weigh clinical, patent, pipeline and safety evidence in one answer?
✓
Built in
Trials, patents, FDA labels, FAERS and Pipeline Graph's view of who is developing what, in the same answer as the biology
◐If connectors are set up
◐Clinical and patent data described for preclinical work; safety data not identified in public sources
Does every answer check its evidence and show what it did not cover?
✓
Coverage estimate in every answer
Each answer estimates how much of the relevant evidence it found and names the parts of the question the evidence did not reach, alongside a limitations section
◐A reviewer agent checks citations and figures; a coverage estimate was not identified in public sources
◐A validation layer screens out weakly supported claims; a coverage estimate was not identified in public sources
Does it reason over a curated base of known relationships?
✓
Curated knowledge graph
Reasons over a built in curated knowledge graph that is updated every two weeks, and also synthesizes information from connected sources
—Uses scientific skills and connectors to query and synthesize information from connected sources
✓Reasons over a prebuilt knowledge graph focused on preclinical biology
Can a committee audit the method?
✓
Same workflow, every run
Scientific Workflows encode a team's steps and criteria and keep an auditable record of each run
✓Auditable output histories, with preconfigured and custom skills
✓Prebuilt research workflows, with a compliance trail for each run
✓Built in◐Partial or depends on setup—Not identified in the public sources reviewed

Based on public product information reviewed on October 29, 2026. Product names are trademarks of their respective owners.

What R&D work does Causaly support?

Causaly supports the R&D work teams have to defend before committing resources, from assessing a target and prioritizing an indication to reading a safety signal, tracking competitors and finding biomarkers. Each team can apply its own assessment method, so every review follows a standard it can audit.

Assess a target

Weigh genetic, mechanistic, safety and translational evidence before a target enters the portfolio.

Example question

Assess the human genetic, mechanistic and safety evidence for NLRP3 in ulcerative colitis

Prioritize indications

Compare candidate indications on mechanistic fit, depth of evidence and competitive density.

Example question

Rank inflammatory indications for a TYK2 inhibitor by mechanistic fit and competitive density

Interpret a safety signal

Separate class effects from compound-specific risk across trials, case reports and adverse event data.

Example question

Which adverse events reported for JAK inhibitors have mechanistic support, and which are confounded?

Map a competitive landscape

See every program against a target by modality, phase and sponsor, refreshed as pipelines change.

Example question

Map KRAS G12D programs by modality, phase and sponsor

Find candidate biomarkers

Surface biomarkers with directional evidence for patient selection or treatment response.

Example question

Which circulating biomarkers have evidence for predicting response to anti-fibrotic therapy in IPF?

Monitor a program

Track new trials, publications, patents and safety reports for a mechanism without rebuilding the search.

Example question

Show what changed this quarter for programs that share our lead asset's mechanism

How are R&D teams using Causaly?

Scientists across discovery, translational research and clinical development describe what changed in their work.

Target validation

“Companies live and die based on the identification and veracity of their targets. If you make a mistake at the very first step, every step that follows will be a snowball effect from that initial wrong decision.”

Dr. Siddhartha Mukherjee
Co-founder, Manas AI
Read the case study →
Early discovery

“I use it daily for early ideation, hypothesis validation, biomarker exploration, safety signal monitoring. It’s become the starting point for the kind of deep scientific questions that used to take days to get a foothold on.”

Eisai
Scientist, Early Discovery and Translational Research
Read the case study →
Clinical program evaluation

“If it outputs a table, I can pull the exact references behind each point. It’s faster and reduces the risk of missing key perspectives.”

Top 10 pharma company
External Innovation Advisor
Read the case study →

Frequently asked questions about Causaly

What is Causaly?

Causaly is an AI platform for life sciences R&D that applies a team's own methods and evidence standards to scientific questions, producing work that scientists can check, question and act on.

Does Causaly work with external AI tools and models?

Yes. Causaly works inside Microsoft 365 Copilot and connects to other assistants and agent platforms through MCP, A2A and APIs.

Which data sources does Causaly use?

Causaly draws on the published literature, licensed full text, clinical trials, patents, genetic studies and curated biomedical databases, together with your organization's own data where it is connected.

How current and reliable is the data?

Causaly's knowledge graph is refreshed every two weeks, and news and web sources are searched at the moment each question is asked. Every citation is verified against its source before an answer is shown.

Can my organization's internal data be connected to Causaly?

Yes. Scientists can analyze their own documents alongside the published evidence, and organizations can connect internal tools or bring their data into a dedicated, isolated Causaly environment.

How should biotech teams structure R&D data for AI?

Give every gene, compound and disease a consistent identifier, record where each result came from and keep the experimental context with the data. Causaly maps internal and public data to the same entities, so both can be weighed together.

How does Causaly pricing work?

Pricing depends on the number of users and the capabilities your teams need. Contact the Causaly team for a quote.

Does Causaly train models on customer data?

No. Customer data is never used to train AI models, and Causaly is ISO/IEC 27001 certified.

See Causaly work through your hardest open question

Bring a target, indication or safety signal your team is weighing now. We’ll run it through Causaly with you and show the evidence behind the answer, what contradicts it and what is still missing.

Try Causaly
Causaly logo

Unlock scientific insights from external and internal information at unprecedented levels of precision and efficiency.

Get started
Login
Causaly
AboutRequest a demoContact usResourcesCareers
LEGAL
Privacy policy
Terms of service
Cookies settings
© Causaly. All rights reserved.