AI assurance platform for major project delivery

Assess any document set against any criteria framework.

Built for major infrastructure programmes. Every finding cited to the exact paragraph, with human in the loop.

Your methodology, our engine.
programme-insights.app / assessment-4127
Live
Outline Business Case — North Hub
87 pages · DOCX · IPA Gateway 2
3.2 Strategic Case
4.1 Benefits Register
5.4 Financial Case — Optimism Bias
6.2 Commercial Viability
Findings
0 of 4
0
Criteria in IPA Gateway module
£0M+
Programme scale we serve
0+
Live assessment modules
0
Steps per criterion · every finding traceable
Agentic pipeline

Not a single AI call. Eight steps. Every one traceable.

Every assessment criterion runs through an eight-step pipeline designed for auditability — not speed. Retrieval, verification, deterministic scoring, and self-critique. The extended tier adds multi-persona debate and cross-criterion validation for the highest-stakes decisions.

1 Step one

Hybrid retrieval.

Dense vector search and BM25 keyword search, run in parallel across your document set. Every passage scored for relevance — nothing buried by embedding noise alone.

Retrieval
2 Step two

Retrieval quality check.

Before reasoning begins, the system grades its own retrieval. If the top passages don't meet a confidence threshold, it re-searches with reformulated queries. No assessments built on thin evidence.

Self-correcting
3 Step three

Binary decomposition.

Complex criteria are broken into specific yes/no sub-questions. "Is the risk register versioned?" "Does it include residual risk?" Each decided independently against the evidence.

Decomposition
4 Step four

Deterministic scoring.

Ratings are calculated from the sub-question results using a defined rubric. The same evidence produces the same rating every time. No temperature, no improvisation.

Deterministic
5 Step five

Citation verification.

Every quoted passage is re-read against the source document. Citations that don't match verbatim are rejected and the reasoning is re-run. Hallucinated citations never survive.

Verified
6 Step six

Self-critique.

The pipeline reviews its own output against the rubric and flags weak reasoning, missing evidence, or logical gaps. Borderline ratings trigger a second pass.

Conditional
7 Step seven

Multi-persona debate.

Three AI specialists — each with a distinct role — independently assess the same evidence. Disagreements are surfaced to a human reviewer. No consensus-by-default.

Extended tier
8 Step eight

Cross-criterion validation.

The final pass checks for contradictions across the full assessment. If risk is red but the financial case calls it manageable, the system flags it. Consistency isn't optional.

Extended tier
Per criterion · fully traceable
8-Step agentic assessment
Watch the pipeline progress as you scroll. Each step produces auditable artefacts.
Step output
Human in the loop

AI assesses. Your team decides.

Every finding can be accepted, challenged with context, or force-overridden — with mandatory justification. Each action writes to a tamper-evident audit trail that flows directly into board-ready reports. Try it below: the panel on the left is live.

Review queue · IPA Gateway 2
North Hub OBC · 14 findings · 3 awaiting review
AMBER — gaps identified
Criterion BC-14 · Financial case
Optimism bias adjustment applied at 15% — below HMT Green Book guidance of 40% for standard infrastructure at OBC stage.
Amber
Source document applies a flat 15% bias across all cost lines. Green Book Annex A.1 requires optimism bias of 40% for standard civil works at outline business case, reducing to 22% post-risk-adjustment. No mitigating justification for the lower value is evidenced in sections 5.4–5.7.
North Hub OBC · §5.4 ¶3 · p. 42
Quick prompts →
Finding accepted
Logged to audit trail. Moving to next in queue.
Evidence mapping

How a finding gets cited.

Every finding traces to a specific paragraph in a specific document. Scroll to watch the chain form: finding → source line → quoted text → confidence score. This is the chain that goes into your board report.

1 Finding raised

The pipeline issues a finding.

Against criterion BC-14 — financial case — the scoring step returns amber. The finding card renders on the reviewer's queue with the criterion reference, rating, and short rationale.

2 Citation line

A line draws back to the source.

The system knows exactly which paragraph the finding is grounded in. Hovering or clicking the finding draws a line back to the cited passage — no hunting through the document.

3 Paragraph highlights

The source text highlights.

The cited paragraph is brought into focus and the quoted span is surfaced in context. You see the exact phrase the system relied on — not a paraphrase, not a summary.

4 Confidence exposed

Confidence becomes visible.

Alongside the rating sits a confidence score derived from retrieval quality and sub-question agreement. Low confidence is a signal — not a verdict. You decide what to do next.

North Hub OBC · §5.4
5.4.1 Cost assumptions
5.4.2 Optimism bias
"A flat optimism bias of 15% has been applied across all cost centres."
5.4.3 Risk provision
5.4.4 Sensitivity
Finding — BC-14
Financial case · Optimism bias
Optimism bias applied at 15% — below Green Book guidance of 40% at OBC.
AmberVerified cite
Confidence
0%
High retrieval quality · sub-questions aligned
Content generation

Remediation drafts — with provenance baked in.

When a finding needs action, the platform drafts the remediation content — section text, risk entries, mitigations — grounded in your own documents. Every generated field carries a confidence score and a citation chain. Hover any field to see where it came from.

Consultants edit in place, approve, and push to the source — no copy-paste between tools, no lost lineage.

  • Field-level confidence — see what the AI is sure of and what it isn't.
  • Every inserted fact carries a citation to a specific paragraph.
  • Track-changes view before you commit back to SharePoint.
Remediation draft · BC-14
Proposed addendum — §5.4 Optimism bias

Following review, optimism bias will be revised to 40%Green Book A.1 · p. 22 · 94% at outline stage, in line with HMT Green Book Annex A.1 for standard civil infrastructure. This will reduce to 22%Green Book A.1 · post-QRA · 91% following completion of the quantitative risk assessment scheduled for Q3 2026Programme plan · §3.2 · 87%.

Supporting analysis — including the P80 confidence intervalRisk register · §7.2 · 93% — is attached in Appendix C. All figures reconcile to the base cost estimate of £127.4MCost plan v4 · §2.1 · 96%.

Drafted 14:22 · 5 fields · 4 citations 92% avg confidence
Board-ready output

Reports your board can read — with citations your auditor can check.

Export the full assessment to PDF or DOCX with one click. Every rating carries its rationale, its citations, and the full audit trail of human decisions. Nothing decorative. Everything defensible.

  • PDF and DOCX — both preserve the full citation chain.
  • Appendix per criterion — rationale, sub-questions, evidence, decisions.
  • Review-ready formatting — no reformat before it goes to the SRO.
IPA Gateway 2 Assessment — Export
IPA Gateway 2 · North Hub programme
Generated 19 Apr 2026 · 236 criteria · 14 findings
Amber
BC-14 — Optimism bias below Green Book guidance
§5.4 ¶3 · p. 42 · verified cite
Amber 92% conf.
BC-07 — Benefits register lacks owner per benefit
§4.1 ¶2 · p. 28 · verified cite
Red 96% conf.
BC-02 — Strategic fit with national policy
§3.2 ¶1 · p. 14 · verified cite
Green 89% conf.
Working with your documents

Connected once. Current always.

SharePoint libraries are monitored for change — and answered through a chat interface that cites every sentence back to the source.

SharePoint integration

Point the platform at a library. It indexes everything, monitors for change, and re-assesses only what's affected when a document updates.

North Hub · OBC pack
12 docs · 847 pages · last sync 2h ago
Live
Corridor A · risk library
4 docs · 162 pages · last sync 18m ago
Live
Financial case updated — 3 criteria re-assessing
Change detected 14 minutes ago. Only BC-14, BC-15, and BC-18 are being re-run. Your readiness score updates when complete.

Chat with your documents

Ask a question across the full document set. Every answer lands with paragraph-level citations you can click through to.

MD
What's the current optimism bias position and where is it justified?
PI
Optimism bias is applied at 15% across all cost centres§5.4. No explicit justification is given in the financial case; it references the risk register§7.2 but the register doesn't contain a mitigating rationale. Green Book Annex A.1 expects 40% at OBC for standard civils.
MD
Is there a QRA that would change that?
PI
Programme plan schedules a quantitative risk assessment in Q3 2026§3.2. It has not yet been completed — the current 15% figure pre-dates it.
Ask across the document set…

See it running on your review pack.

A 30-minute discovery session. We'll assess one of your documents against the framework of your choice — so you can see what the platform finds, not just how it looks.