Method
The method shows how AI Advisor moves from decision context and material to analysis, risk assessment and decision-ready reporting.
Which decision?
The analysis starts from the decision the review is meant to support. This may concern investment, procurement, contracts, AI adoption, a transaction, a board decision or another risk-sensitive situation. The review is scoped based on the decision’s risk level, consequences, timeline and available material.
Which claims?
The claims about technology, AI, data, security and rights that need to be tested before the decision are identified. This may include claims about proprietary software, IP control, AI functionality, data sources, security level, integrations or dependencies on vendors and third-party services.
What can be verified?
Contracts, system documentation, architecture sketches, security material, vendor lists, interviews and, where needed, technical review are used to test what the technology consists of, who controls it and which dependencies affect the decision.
Analysis
Findings are assessed by risk, exposure, materiality and confidence in the information on which the assessment is based. Information is classified as verified, partly verified, unverified, contradictory or not assessable. This makes uncertainty visible before it is built into price, terms, contracts, risk acceptance or board decisions.
Findings are translated into decision points
The final decision material states the relevance of the findings and their possible impact on terms, warranties, vendor approval, disclosures, risk acceptance, remedial actions or the need for deeper review. If the review relates to an investment or transaction, the report may also identify issues that should be considered in valuation, price discussions or continued commercial analysis.
The decision material can be used by lawyers, management, boards, investment teams or advisers. It is written to make technical and legal risks understandable without simplifying away their significance.
Consistent working method
To create consistent, traceable and comparable analysis, AI Advisor uses the internal analysis and documentation tool Epistera. The tool supports structured information gathering, risk indicators, confidence assessment of material, uncertainty markers, escalations, traceability and report preparation.
The client value does not lie in the tool itself, but in the integrated review and decision-ready material. Epistera strengthens the working method by making the analysis more consistent, documented and possible to follow afterwards.
Technical specialist when needed
The initial review identifies which technical questions are material to the decision and whether the available material is sufficient for a reliable assessment. When deeper technical verification is required, specialist competence may be included, for example to assess architecture, security, code quality, open source, AI/ML dependencies or transaction-specific technical analysis.
Confidentiality and material handling
For this type of engagement, confidentiality and careful handling of material are fundamental. Confidentiality covers everything entrusted to AI Advisor within the scope of the business or that AI Advisor becomes aware of in connection with the engagement. Exceptions apply only where consent is given or where there is a legal obligation to disclose information.
If a technical specialist or other subcontractor is involved, this occurs only where the scope of the engagement requires it. Such party is subject to corresponding confidentiality obligations and receives only the material required for the scoped review.