Pilot Demo Report · Confidential
GrundMind AI Adoption Fit Diagnostic

Where AI creates value,
where value leaks,
and what to redesign next.

A pilot diagnostic of AI adoption fit at Solvexa — assessing value capture, workflow integration, cognitive alignment, and the operating practices that turn AI usage into measurable business performance.
Illustrative demo — fictional data shown for design purposes only
Prepared for
Solvexa d.o.o.
Lena Hartmann & David Renner
Ljubljana, Slovenia
Prepared by
Grund Institute
GrundMind AI Adoption Practice
Switzerland
Pilot edition
v1.0 · Demo
Sample report — illustrative use only
Not for distribution as actual findings
§ 01 — Executive Summary
page · two
For the Board of Solvexa

A board-ready view of where AI is, and is not, creating value.

AI usage has begun across Solvexa, but value capture is uneven. This summary reflects a fictional pilot illustrating how GrundMind diagnoses the gap between AI activity and AI-driven business outcomes.
Finding · I

AI usage exists, but is uneven across teams

Several teams use AI tools daily; others use them rarely or informally. There is no shared definition of what "good" AI use looks like, and no shared measure of whether it is helping the business.

Finding · II

AI is used for individual productivity, not workflow transformation

The dominant use case is personal task assistance — drafting, summarising, searching. AI is rarely embedded into the workflows where Solvexa earns or loses the most time: implementation, support, and documentation.

Finding · III

Trust and verification practices are inconsistent

Some teams verify AI output rigorously, others accept it at face value. Verification is informal, not standardised, and rarely captured as a reusable practice. This creates both quality and risk variability.

Finding · IV

The biggest opportunity is workflow redesign, not more tools

The highest-value opportunity is redesigning implementation, support, documentation, and internal knowledge workflows around AI — reducing rework, shortening response times, and lifting documentation quality.

Executive Conclusion

Solvexa should not focus first on buying more AI tools. The leverage point is redesigning the workflows where AI can reduce rework, improve documentation quality, shorten response times, and support better decision-making.

§ 02 — AI Adoption Fit Scorecard
page · three
Illustrative demo data
The Four Dimensions of AI Adoption Fit

A composite view of fit, not adoption activity.

GrundMind measures AI Adoption Fit across four dimensions. Each is scored 0 – 100. Together they describe whether AI activity is converting into business value — and where the conversion is breaking down.
Solvexa · Composite Diagnostic

Overall AI Adoption Fit

Weighted across the four GrundMind dimensions

Composite Score
52/100
Activity present · Conversion limited
62 VALUE FIT

Value Fit

Activity present, conversion partial
What it meansAI use is connected to some valuable activities, but value capture is not yet measured or systematic.
Why it mattersAI activity without value conversion is a cost, not a return.
Leadership implicationIdentify 2–3 priority value pools where AI must demonstrably move a KPI.
41 WORKFLOW FIT

Workflow Fit

Lowest score · Highest leverage
What it meansAI lives next to workflows, not inside them. Most use is ad-hoc, individual, and disconnected from end-to-end processes.
Why it mattersWorkflow embedding is where rework, cycle time, and quality gains compound.
Leadership implicationSelect one priority workflow and redesign it end-to-end with AI as an embedded layer.
58 COGNITIVE FIT

Cognitive Fit

Diverse styles, no shared model
What it meansSolvexa employees engage with AI in distinctly different cognitive modes. There is no shared model of how the team should think with AI.
Why it mattersCognitive misalignment produces silent drag — friction that is real but unmeasured.
Leadership implicationTailor AI enablement by interaction style, not by generic prompt training.
46 OPERATING FIT

Operating Fit

Light governance, soft standards
What it meansAI usage operates without verification rules, shared KPIs, or quality standards. Experimentation is high; consistency is low.
Why it mattersScaling AI without operating discipline is how organisations introduce risk faster than value.
Leadership implicationDefine lightweight verification rules and connect AI usage to existing operational KPIs.
— interpretation —

Read the pattern

A "scattered start" profile.

Value Fit (62) is the strongest dimension because AI activity is real and pointed at relevant tasks. Workflow Fit (41) is the weakest — AI is not yet a feature of how Solvexa operates. This is the most common starting profile for early-stage AI adoption in B2B software companies.

What 52/100 implies

There is upside, not a problem.

A composite of 52 means Solvexa has crossed the experimentation threshold but has not yet entered systematic value capture. The score gap between Value Fit and Workflow Fit (21 points) is the actionable signal: capability exists, conversion is missing.

§ 03 — Value Leakage Estimate
page · four
Illustrative model · Not a confirmed financial result
Where time, effort, and quality may be leaking

An annualised view of hidden cost from AI–workflow friction.

Value leakage is the time, rework, and quality drag that is created when AI sits outside of workflows. The model below is illustrative — not audited — and shows the order of magnitude that even small per-employee frictions can produce.
Illustrative calculation

Hidden Leakage Model

Fictional baseline assumptions, Solvexa pilot demo.

Affected employees 80
Lost minutes per person per day 10
Working days per year 220
Fully loaded labour rate €35/hr
Estimated annual leakage
~ €102,667
Illustrative model only. Not a confirmed financial result. Actual figures depend on team composition, ERP cycle profile, and verification practices at Solvexa.

Leakage Sources

Where the friction is most likely concentrated, ranked by estimated contribution.

AI output requiring heavy rewriting 28%
Duplicated verification across teams 22%
Inconsistent AI use across roles 18%
Poor integration into ERP implementation 14%
Informal use by support teams 10%
Manual, fragmented documentation 8%

Read the leakage carefully

The €102,667 figure is illustrative. The more important number is the shape of the leakage — concentrated in rework, duplicated verification, and inconsistent practices. These are workflow problems, not tool problems. Buying more AI tools will not address them. Redesigning a single high-leverage workflow plausibly will.

§ 04 — Workflow Gap Map
page · five
Illustrative · Four representative Solvexa workflows
Where AI is, and is not, embedded

Four workflows. Four current gaps. Four redesign opportunities.

The four workflows below represent core Solvexa value chains where AI integration is either weakest or most consequential. Each is mapped against its current use, gap, opportunity, and recommended redesign.
Workflow
Current AI Use
Current Gap
Value Opportunity
Redesign Recommendation
ERP implementation documentation
Ad-hoc, consultant-led; AI drafts used inconsistently.
High GapNo shared template, no quality standard, heavy rewriting.
Faster, more consistent implementation deliverables; reduced consultant rework hours.
Introduce a shared AI-assisted documentation template with verification checkpoints embedded in the delivery method.
Customer support response preparation
Informal personal use; not visible in ticket system.
High GapQuality of AI-assisted replies depends on individual skill, not process.
Shorter first-response time; more consistent voice; fewer escalations.
Build an AI assist layer inside the support workflow with response templates, verification rules, and KPI tracking.
Internal knowledge search
Some employees use AI; many still ask colleagues or search manually.
Medium GapKnowledge is fragmented across drives, tickets, and tribal memory.
Reduced knowledge-search time; better onboarding; lower expert dependency.
Centralise the knowledge base and pair it with an AI retrieval layer accessible from existing tools.
Product / development requirement clarification
Light AI use in analysis; almost none in clarification dialogue.
Medium GapAmbiguous requirements still resolved through long meetings.
Faster requirement clarification; fewer rework cycles; cleaner specifications.
Use AI to structure pre-meeting clarification briefs and surface inconsistencies before development.
Workflow × Dimension Heatmap

Where each workflow is strongest and weakest.

Higher copper intensity = larger gap. Each cell is a 1–5 illustrative rating of gap severity.

Workflow
Value Fit
Workflow Fit
Cognitive Fit
Operating Fit
ERP implementation docs
3
5
3
4
Customer support prep
2
4
4
5
Internal knowledge search
3
3
2
3
Requirement clarification
3
3
2
2
Low gap High gap
§ 05 — Cognitive Fit Profile
page · six
Illustrative · Four GrundMind interaction archetypes
How people at Solvexa interact with AI

Four interaction styles. None better. All different.

The GrundMind archetypes describe distinct cognitive interaction patterns — how individuals think with probabilistic systems. They are not personality types and not performance ratings. Each style has strengths, blind spots, and a distinct enablement need.
Co-thinkers
Explore and ideate with AI. Use it as a thinking partner, not a search engine.
Calibrators
Verify and refine AI output. Treat AI as draft material to improve and validate.
Controllers
Need structure, governance, and control points. Trust AI when behaviour is bounded.
AI-averse users
Avoid AI if it adds cognitive friction. Often skilled in deterministic, high-expertise work.
28%
34%
24%
14%
Co-thinkers Calibrators Controllers AI-averse
Co-thinkers
Trust patternHigh, often early.
VerificationLight; prefers iteration over checking.
Risk patternOver-rely; under-document.
Support needGuardrails, not gates.
Calibrators
Trust patternConditional; rises with evidence.
VerificationHabitual; structured.
Risk patternSlow adoption; bottleneck risk.
Support needVerification rules + standards.
Controllers
Trust patternProcedural; tied to structure.
VerificationFormal checkpoints required.
Risk patternAdoption gap if no governance.
Support needBounded use cases + KPIs.
AI-averse users
Trust patternLow by default; protects expertise.
VerificationPrefers traditional sources.
Risk patternSilent disengagement.
Support needIndirect AI exposure; expert-led.

Distribution reading

Solvexa is "Calibrator-heavy".

Roughly two-thirds of employees (Calibrators + Controllers = 58%) require structure and verification practices to engage confidently with AI. Co-thinkers (28%) are likely the visible AI champions; AI-averse users (14%) are at risk of silent disengagement.

What this implies

Generic prompt training will not move the majority.

A Calibrator-heavy organisation needs verification rules and quality standards before it needs more prompt training. The fastest way to expand AI adoption at Solvexa is to make AI use feel structured and bounded — not freer.

§ 06 — Priority Findings
page · seven
Top Seven · Ranked by impact

The findings that matter most for the next 90 days.

Ranked by the combination of business impact and ease of action. Each finding is paired with a business implication and a recommended action that Solvexa leadership can sponsor directly.
01
AI is used more for individual task help than workflow-level transformation.
Business implication

Productivity is improving in pockets, but the company-level outcomes (cycle time, rework, customer response) are not yet moving.

Recommended action

Move from "AI per person" to "AI per workflow". Pick one workflow and instrument it.

02
Documentation is a high-value use case but lacks a shared standard.
Business implication

Implementation documentation quality varies by consultant. AI helps speed, not consistency, without a standard.

Recommended action

Introduce an AI-assisted documentation template owned by the delivery practice.

03
Support and consulting teams may duplicate effort when verifying AI output.
Business implication

Hidden verification work is one of the largest sources of estimated leakage in the model.

Recommended action

Define which role verifies which AI output, and where verification is recorded.

04
Trust in AI output varies by role and experience level.
Business implication

Without calibrated trust, some teams over-use AI and others under-use it. Both reduce value.

Recommended action

Establish a small set of "trust levels" tied to use cases and verification practices.

05
Some users over-rely on AI for speed; others avoid it due to uncertainty.
Business implication

Two opposite failure modes coexist in the same company. They cannot be fixed with the same intervention.

Recommended action

Tailor enablement by interaction archetype. Provide guardrails to Co-thinkers, structure to Controllers.

06
AI use is not yet connected to clear KPIs.
Business implication

Without KPIs (cycle time, quality, rework, response speed), AI usage cannot be evaluated or defended.

Recommended action

For the selected pilot workflow, define 2–3 KPIs before redesign begins.

07
Governance is light, which enables experimentation but creates inconsistency.
Business implication

Solvexa still benefits from low-friction experimentation, but inconsistency will become a brake at scale.

Recommended action

Introduce a "thin governance" layer: verification rules, escalation patterns, KPI ownership.

§ 07 — 30 / 60 / 90 Day Intervention Plan
page · eight
From diagnosis to action

A focused 90-day path from activity to measurable value.

The plan below is deliberately narrow. Its purpose is to demonstrate measurable workflow value within one quarter, before broader rollout. Each phase has a clear exit condition.
— Phase 01 —

30 days

Frame & focus
  • Define 3 priority AI use cases tied to business outcomes.
  • Create verification rules for AI-assisted output.
  • Identify AI champions and calibrators across teams.
  • Select one workflow for redesign — highest leverage, lowest risk.
— Phase 02 —

60 days

Redesign & embed
  • Redesign the selected workflow with AI embedded, not bolted on.
  • Create an AI-assisted documentation template for that workflow.
  • Define 2–3 KPIs (cycle time, rework rate, response time).
  • Train teams by use case, not by generic prompts.
— Phase 03 —

90 days

Measure & scale
  • Measure impact against baseline; share results with the leadership team.
  • Scale the redesigned pattern to a second workflow.
  • Formalise the thin governance layer that emerged from the pilot.
  • Build a leadership dashboard for AI Adoption Fit at Solvexa.

Exit condition for the 90 days

Solvexa leadership should be able to point to one redesigned workflow with measurable value, a thin governance layer that emerged from the work, and a credible second workflow ready for the next sprint. If those three are present, scaling is a business decision, not a question of feasibility.

§ 08 — Executive Readout
page · nine
Board-ready summary

The whole report,
on one page.

GrundMind · AI Adoption Fit
Solvexa Pilot Demo · v1.0
Overall Diagnosis
Solvexa has early AI adoption activity, but value is not yet systematically captured.
Biggest Opportunity
Embed AI into documentation, support, and implementation workflows where rework and knowledge friction are highest.
Biggest Risk
Scaling AI usage without workflow standards, verification rules, and clear KPIs.
Composite Score
Overall AI Adoption Fit at Solvexa sits at 52/100 — activity present, conversion limited.
Recommended Next Decision
Run a focused 60-day workflow redesign sprint for one high-value AI use case.
›

This page is the only one a busy executive needs to read. Every other section in this report exists to support, justify, or operationalise the four statements above.