Research

What is Research?

A literature summary that reads smoothly is usually hiding the disagreements. Research builds an evidence graph instead: what has been claimed, by whom, supported by what, and where the field disagrees with itself. Patents and datasets sit alongside publications, and reproducibility signals are surfaced rather than assumed — so the output supports a judgement rather than substituting for one.

Problems it solves

• Synthesise a body of literature on a question
• See where the field genuinely disagrees
• Include patents and datasets alongside publications
• Surface reproducibility signals
• Keep every claim attached to its source

Key capabilities

Literature Synthesis

Gather published work on a question and separate what is established from what is contested.

Evidence Trail

Every figure carries the record it came from, so a reader can check the answer instead of trusting it.

Fact / Inference Separation

Keep what is known, what is inferred and what is alleged visibly distinct.

Confidence Ceilings

Confidence is capped by the strength of the underlying basis, so a weak source cannot produce a strong-sounding claim.

Agentic Investigation

Specialist agents pursue a question through several controlled steps rather than answering in one pass.

How it is built

Evidence graph

Claims, sources and support are a structure, not a paragraph.

Disagreement preserved

Contested findings stay contested in the output.

Beyond publications

Patents and datasets are part of the landscape.

How it works

How the work is structured, and what each step's output rests on.

1. 1

   Frame the question

   Source

   State what is being asked precisely.

2. 2

   Gather sources

   Source

   Collect publications, patents and datasets.

3. 3

   Extract claims

   Observed

   Record what each source actually claims.

4. 4

   Build the graph

   AI-inferred

   Connect claims, support and contradiction.

5. 5

   Report with disagreement intact

   AI-inferred

   Separate established from contested.

Example use cases

Each example has its own page you can read and share.

Ask Research

Illustrative examples. Results depend on your own data.

• “What is the state of evidence on this question?”
• “Where does the literature disagree?”
• “Which findings have been reproduced?”
• “What does the patent landscape look like here?”

Security, governance and human control

Where the software deliberately limits itself.

Questions

Does it write my literature review?

It builds the evidence structure and keeps disagreement visible. The argument is yours to make.