See how Taltru turns evidence into a traceable human decision.
A 20-minute, fictional-data demonstration followed by a discussion of one evidence workflow. Initial validation focuses on principal-contractor workflows in England.
Research prototype in active validation | no real evidence accepted through the websiteThe workflow, not the sales pitch.
The discussion examines how source records, project requirements, unknowns and authorised decisions currently connect. The aim is to determine whether Taltru addresses a real operational problem and whether a bounded Diagnostic or pilot merits further work.
A working mechanism with explicit limitations.
The prototype accepts fictional text or sample records, identifies relevant fields and evaluates them against deterministic demonstration rules. Automated extraction may be available when the hosted function is configured; otherwise the interface visibly labels its pattern-matching fallback. Neither method makes the final decision.
Fictional decision overview showing supported findings, blockers, unknowns and decision state.
What the current prototype demonstrates.
Deterministic Rule Evaluation
AI assists with extraction; a fixed rule engine produces a supported, blocked or unknown result, while authorised people retain the final decision.
Evidence Status Summary
Supported items, blockers and unknowns are shown separately, with every finding traceable to a specific rule.
Decision History
The prototype demonstrates a hash-linked record of automated checks and human decisions. Production tamper-evidence remains subject to server-side implementation and security review.
Human Sign-Off
A named, competent person must explicitly review and record the decision even when every configured demonstration rule is supported.
Decision history: the demonstration records each check and accountable human decision.
Built around real construction roles.
- Principal contractors: examine whether a more explicit evidence and decision trail could reduce uncertainty before mobilisation.
- Health, safety and compliance managers: challenge the rule logic, human-review boundary and change-trigger assumptions.
- Supply-chain and commercial teams: assess where evidence requirements, ownership and escalation currently become unclear.
A few things people usually ask.
Is this available to buy today?
The production platform is not for sale. The separately scoped Evidence Workflow Diagnostic is available now. Managed Evidence Review delivery remains conditional on security, contracting, data-handling and reviewer controls.
Does the AI decide whether we're compliant?
No. AI is used only to assist with extraction. A separate deterministic rule engine evaluates the available facts, and authorised people retain the final decision.
What does Taltru currently cover?
The fictional prototype demonstrates eight rules across insurance, RAMS and CSCS evidence. These are demonstration rules, not a complete statement of law, regulation or customer requirements.
How long does a walkthrough take?
About 20 minutes: a look at the prototype, a discussion of your workflow, and an honest assessment of whether a Diagnostic or pilot conversation is appropriate.
Book a 20-minute walkthrough.
Tell us a bit about your organisation and we'll reach out to arrange a time. There is no commitment created by requesting a walkthrough.
Prefer email directly? Write to info@taltru.com.