01
Agentic collection
Collection agents plan and run the groundwork across connected registries, filings, public records and client documents, recording the source and retrieval date as the material arrives.
§B2B due diligence platform
Due diligence you can trace back to the evidence.
AnKre helps organisations establish which professional and corporate claims they can rely on. Specialist AI agents build a knowledge graph of the evidence, and every finding keeps a traceable connection to the record that supports it — and to the questions that remain unresolved.
Hover or select a node to isolate its relationships and read the evidence attached to it — node type, source, retrieval date, status and the agent that wrote it.
01What AnKre does
It brings together information from corporate registries, regulatory filings, public records and client-provided documents, then turns that material into evidence a reviewer can stand behind.
AnKre supports investigations involving business relationships, investment transactions, institutional partnerships, professional credentials and access to sensitive information. It reduces the manual work of gathering, organising and checking information, so professionals spend their time on judgement rather than collection. The groundwork is carried by specialist analysis agents working over a shared knowledge graph; the judgement stays with the reviewer.
02Core capabilities
02
Structured evidence
Corporate registries, regulatory filings, public records and client-provided documents are organised into structured evidence rather than a folder of attachments.
03
Knowledge graph
Subjects, relationships, claims and source documents are held in a single knowledge graph, so evidence is connected rather than filed — and any node can be followed back to the record behind it.
04
Identity resolution
The same person or organisation appearing across different sources is resolved into a single subject, so findings are attached to the right party.
05
Relationship mapping
Connections between people and organisations are mapped, making ownership, control and association visible across the material gathered.
06
Claim verification
Stated claims are checked against the available records, and the outcome is recorded with the evidence that supports or fails to support it.
03What a finding looks like
Supported, unresolved or contradicted — stated plainly.
Findings are not delivered as a narrative you have to take on trust. Each one is recorded with the record behind it, where it came from and when it was obtained, so a reviewer, a committee or a regulator can retrace the same path.
Security & buyer questionsReport excerpt — findings table
Illustrative example
04The agents behind the work
Specialist agents do the groundwork. A reviewer does the judging.
Each agent has one job and writes what it finds into the shared knowledge graph, with the source and retrieval date attached. Nothing an agent produces enters a report as settled fact — it arrives at a reviewer as a proposal to accept, question or reject.
Collection agent
Plans and runs the search across connected registries, filings and public records, recording where each item came from and when.
Document agent
Reads client-provided and retrieved documents, extracting the statements and figures that matter to the case as citable evidence.
Identity agent
Reconciles the different appearances of one person or organisation across sources into a single subject in the graph.
Relationship agent
Draws ownership, control and association edges between subjects, each edge carrying the record it was derived from.
Claim-checking agent
Tests stated claims against the graph and marks each one supported, unresolved or contradicted, naming the evidence gap where one exists.
Drafting agent
Writes the narrative under an evidence constraint: it can only state what documented findings support, and must name what is missing.
05The traceability chain
Every finding can be reconstructed, link by link.
Collect — The collection agent plans and runs the search across registries, regulatory filings, public records and client documents, recording the source as the material arrives.
01
Claim
What is asserted
02
Evidence
What was found
03
Source
Where it came from
04
Retrieval date
When it was obtained
05
Confidence
How strongly it holds
06
Reviewer action
Who accepted it
AnKre also makes uncertainty explicit. Verified claims are distinguished from unresolved questions, missing evidence is identified, and the limits of the investigation are recorded alongside the findings.
Read how traceability works06What you receive
Executive notes
A short read for decision-makers, stating what the evidence supports and what remains open.
Due diligence reports
The full investigation write-up, with findings tied to the records behind them.
Evidence packs
The underlying documents and records, organised so a reviewer can retrace the work.
Audit trails
A record of collection, checks and reviewer decisions across the investigation.
Automated collection and verification are combined with human review. An evidence-constrained drafting layer is designed to keep generated narratives tied to documented findings and clearly identified gaps.
07Deployment
A modular architecture, built to sit where your work has to sit.
The platform supports different data providers and specialist analysis components. Deployment options are being developed for cloud, sovereign infrastructure and isolated environments.
Deployment options08In plain terms
What is evidence-traceable due diligence?
Due diligence where every conclusion carries the claim, the evidence, the source, the retrieval date, a confidence level and the reviewer decision behind it, so any finding can be reconstructed later by a committee, an auditor or a regulator.
What does AnKre do?
It gathers material from corporate registries, regulatory filings, public records and client documents, structures it as evidence, resolves identities across sources, maps relationships and checks stated claims, then puts the result in front of an analyst to accept, question or reject.
Who is it for?
Teams that must verify people and organisations before a decision: business relationships, investment transactions, institutional partnerships, professional credentials and access to sensitive information. AnKre works with organisations worldwide.
What does agentic mean here?
Specialist AI agents plan and run the groundwork (searching connected sources, reading documents, reconciling identities, drawing relationships and testing claims) and record what they could and could not settle. They propose; an analyst decides.
What is the knowledge graph?
The structure the evidence lives in: subjects, the relationships between them, the claims made about them and the source documents behind each one. Every node and edge carries the record and retrieval date it came from, which is what makes a finding retraceable.
How is uncertainty handled?
Explicitly. Findings are recorded as supported, unresolved or contradicted, missing evidence is named, and the limits of the investigation are stated alongside the conclusions.
Discuss an investigation workflow
Tell us about the checks your teams run today. We will show how the same work looks when every finding carries its evidence.