Data infrastructure for policy and political AI.
TACITUS

TACITUS / AI data infrastructure

Policy expertise.
Structured for AI.

We structure sources and team knowledge for policy research and writing.

Inside a policy knowledge pipeline
Drag to rotate · click a point

Capture the exact text.

Keep the passage, source edition and historical wording together.

Historical source · resolution 2254 (2015)

Inspect a record & its source

Complete scene data ↗ · Discuss your workflow ↗

Text occurrences, published vocabularies and authored mappings are distinct. Recorded AI output remains for review; generation-quality gains need evaluation.

The product

Your work.
With its context.

A private workspace for sources, reusable knowledge and work you can inspect.

Registration is open. AI generation is currently paused.

What you can explore →
Example TACITUS App conversation with a policy question, suggested context and library attachments.
Example workspace using public guidance and a synthetic account. No generated answer is shown.

These describe the work the product is being built around. Explore sources, context and review tools now; hosted AI generation is paused.

The technology

Prepare the knowledge
behind the answer.

Sources, methods and output structures stay distinguishable. Select what a task needs and inspect what it uses.

  1. 01

    Data structures

    Who, what, when—and which source? Keep identities, relationships and editions explicit.

    Inspect the structures →
  2. 02

    Context capsules

    The background, terminology and working method a particular question needs.

    Explore capsules →
  3. 03

    Artefacts

    Templates and saved work you can inspect, review and reuse.

    Explore artefacts →
Why structure the context?

Exact source editions, reviewed identity links and explicit methods make context selection inspectable. Repeatable data handling does not make a model's interpretation deterministic or correct.

Saliency means retaining what matters to the question. Efficiency must include preparation and review—not just prompt size. Both benefits need task-specific evaluation.

Technical foundations →Research and evaluation →

Work with TACITUS

A question.
A useful deliverable.

Commission a scoped research engagement. We agree the question, sources, output and review before work begins.

Research synthesis
A decision brief covering findings, disagreements and open questions.
Policy-evidence tracking
A dated register of changes, requirements, evidence gaps and review responsibilities.
Source-preserving briefs
Written analysis with traceable passages, dates and qualifications.

Scope and commercial terms are agreed directly. An inquiry creates no payment.

Sourced map of semiconductor fabrication geography, using a historical 2022 baseline.Sample analysis

Compute diplomacy

Where do alternative suppliers share manufacturing exposure?

2022 fabrication baseline · historical research, not a live capacity feed. Sources and method →

Questions

Before you begin.

All questions and references →
What is available today?

Register to explore the App's source, library and review tools. AI generation and payments are not active. Public code is available separately, and scoped research engagements are agreed through Consulting.

Access and engagements →
Why use structured context?

A policy question often depends on organisational knowledge, a precise source edition or an expert method. TACITUS makes that context explicit and reusable, while keeping evidence separate from instructions and interpretations.

How it works →
How will the benefit be measured?

Through matched tasks, retained failures and independent review of source support, relevance and editing effort. Software tests and academic papers do not establish a product quality or efficiency gain.

Evaluation approach →

Source inspection