AI, acquisitions & divestments
Better evidence.
Clearer company decisions.
A practical approach to acquiring businesses, improving their performance and preparing them for their next owner.
We combine industrial judgement with AI-assisted research and analysis. The aim is to make evidence easier to find, assumptions easier to challenge and decisions easier to explain.
Discuss your company’s next stage
One approach across the ownership cycle
Acquire. Improve. Prepare for exit.
Each engagement starts with the decision the owner or buyer needs to make. We agree the scope, information permissions, review responsibilities and expected outputs before selecting the AI tools.
01 · AcquireTest the investment case
Define the target profile and assess commercial, technical and operational fit. Review customer concentration, product economics, intellectual property, equipment and supply-chain dependencies alongside financial diligence.
AI’s role: organise approved documents, surface inconsistencies and draft source-linked questions. Separate reported results, adjustments and forecast assumptions.
Output: an evidence-backed investment thesis, diligence questions, risk register and integration priorities for the deal team to review.
02 · ImproveTurn the thesis into action
Translate opportunities into a delivery plan with accountable owners, costs, milestones and a measured starting point. In aerospace and manufacturing, assess inspection, maintenance, planning and energy use against real operating constraints.
AI’s role: connect operational data to improvement hypotheses, compare scenarios and track pilot results. Distinguish demonstrated gains from opportunities still to be tested.
Output: a prioritised improvement plan and performance scorecard, including a first-100-days plan where appropriate.
03 · DivestPrepare a credible company story
Assess sale, carve-out and ownership-transition options. Clarify the business perimeter, buyer fit, standalone requirements and the evidence supporting the company’s position.
AI’s role: help organise the data room, draft management questions and identify gaps or contradictions across approved materials.
Output: an exit-readiness assessment, evidence map, reviewed Q&A and separation or transition priorities.
Manufacturing example · Digital twins
Test an inspection approach before investing.
A robotic X-ray inspection twin illustrates how manufacturers can explore an inspection cell virtually. The example simulates part positioning, scan geometry, calibration errors and image reconstruction, helping engineers examine how equipment choices and operating assumptions affect the result.
01 · ModelDefine the inspection problem
Start with the part, the features to inspect and the required image quality. Use a virtual cell to explore robot reach, equipment placement and scan coverage before committing to a physical trial.
02 · CompareExplore the trade-offs
Compare nominal and calibrated geometry, vary scan settings and inspect reconstructed slices against the simulated part. Keep model assumptions and unmodelled constraints visible.
03 · ValidateBuild evidence for deployment
AI could help prioritise scan settings or flag candidate defects for engineering review. Validate these capabilities on representative physical parts, measuring defect detection, false calls, inspection time and repeatability against an agreed baseline.
For owners and acquirers: use this approach to challenge equipment assumptions, scope an operational improvement pilot and identify the evidence needed to support a capital investment.
Scope: this is a simulation example, not a demonstrated production AI deployment or a claim of achieved savings. Simulated reconstruction and calibration do not by themselves establish defect-detection performance or certify equipment safety.
Explore the digital twin example ↗External demonstration · authorised sign-in required.
Product management · Learning & training
Turn learning into better product decisions.
Product managers need to connect customer evidence, strategic choices and business performance. The Product Management Capability Builder illustrates an approach that carries learning beyond a workshop: assess understanding, practise real decisions and return to the gaps over time.
01 · AssessFind the development priorities
Compare self-assessment with responses to practical questions. Use adaptive practice to revisit weaker areas while keeping established skills in rotation. Treat the profile as a guide for development and discussion.
02 · PractiseRehearse the decision
Work through customer evidence, prioritisation, commercial trade-offs and stakeholder influence. The example links to separate AI role-play training for executive updates, alongside structured briefing exercises and self-review.
03 · ApplyBuild a working habit
Translate learning into a short action plan with specific commitments. Return to exercises through spaced practice, apply them to live product decisions and discuss progress with a coach or manager.
Vassal’s application: help product teams define learning priorities, select appropriate AI practice tools and connect development to clearer business cases, stronger customer evidence and better decision follow-through.
How we assess value: review the quality of a team’s evidence, recommendations and completed actions over time. Assessment scores alone do not demonstrate business impact. Adaptive assessment and AI role-play are distinct parts of this approach; much of the example’s guidance is self-directed.
Explore the product management learning example ↗External capability tool · authorised sign-in required. Linked training may require separate organisational access.
How AI supports the work
From documents to decisions.
Build a reliable evidence base+
Catalogue authorised financial, commercial, technical and operational sources. Record document owners, dates, versions and access rights. Link each material finding to a document and page, table or record, and distinguish facts from assumptions.
Ask, compare and challenge+
Use retrieval-assisted AI to find relevant evidence and draft answers. Flag missing information and conflicting sources. Reconcile units and periods, and calculate material figures in controlled models rather than relying on generated prose.
Review answers before reuse+
Assign substantive answers to a responsible reviewer. Maintain an approved Q&A library with supporting evidence, review dates and unresolved issues. Reopen answers when new documents change the evidence.
Carry decisions into delivery+
Turn approved meeting notes and diligence findings into questions, actions and decisions with named owners. Where participants agree to transcription, AI can help organise the discussion for review. Track decisions and their rationale through acquisition, improvement and exit preparation.
Judgement & accountability
People own the decision.
This is Vassal’s engagement approach. Tools and access controls are selected for each assignment; this public website is not a deal data room or document-upload service.
Confidential material belongs in an approved environment with appropriate access and retention controls. AI drafts require review, and financial, legal and technical conclusions remain with the relevant professionals and decision-makers. AI output is not proof of accuracy or a guarantee of transaction value.
We assess usefulness through evidence coverage, unresolved issues, review turnaround and measured operating results. A persuasive narrative should follow the evidence.
Discuss an acquisition or divestment