Make your AI predictable.
Set the new standard for corporate AI. Integrate ValueFilters to ensure every output perfectly reflects your organization’s unique DNA.
The problem no one is solving
Your AI has read the internet. It hasn’t worked at your organization.
It doesn’t know why you always say no to that type of client. It doesn’t know how your best project lead handles an escalation. It doesn’t know that “we prioritize quality” in your organization actually means you’d rather miss a deadline than ship something half-finished.
That knowledge lives in people’s heads. It gets passed on in coffee rooms, absorbed by working alongside someone for three years, and lost the day they leave.
AI governance platforms can monitor and restrict AI. They can’t give it judgment. Training programs can teach people about AI. They can’t teach AI about your organization.
We do.
Feb 2025
EU AI ACT ARTICLE 4 IN FORCE
AI literacy training is now a legal requirement for all organisations deploying AI in the EU, and the clock is running toward August 2026 for full compliance.
AI without context vs. AI with context
Without ValuesFilters
Every employee gets different AI answers to the same question. Your AI suggests things that violate unwritten rules. New hires get advice that experienced staff would never give. You can’t predict what AI will tell your customers. Compliance requires reviewing every AI output. Knowledge walks out the door every time someone leaves.
With ValuesFilters
AI responds consistently — aligned with how your organization actually works. Suggestions reflect real priorities, not generic best practices. New hires get the same quality guidance as a 10-year veteran. You can predict AI behavior because you’ve coded in your values. Compliance is built into the output, not bolted on after. When things change — new laws, new policies, key people leaving — the system updates with them. Your employees and your AI work from the same context.
Four ways to give your AI organizational judgment
VALUESFILTER
ContextPrompt
Same voice from every AI tool. Every time.
VALUESFILTER
ValuesFirst
Explain your AI decisions before anyone asks.
VALUESFILTER
Tacit Knowledge
Keep the expertise when key people leave.
VALUESFILTER
Raise-AI
Real AI literacy — not just a certificate on the wall.
What this takes from your team
4–6 weeks. About 8 hours total from your side. We do the heavy lifting.
Week 1 — Discovery & Kickoff
Your time: 1 hour (project sponsor)
Map your organization’s values, decision culture, and AI usage patterns.
Weeks 2–3 — Deep Interviews
Your time: 3 hours per person interviewed
Structured interviews using Cognitive Task Analysis — the same methodology used in aviation, healthcare, and defense to capture expert judgment under uncertainty.
Weeks 3–4 — Analysis & Validation
Your time: 1 hour feedback session
Cross-validate findings, identify patterns, code knowledge into narrative format.
Weeks 4–5 — Living Book
Your time: 2 hours review
Write your 120-page Living Book — your organization’s judgment, codified.
Weeks 5–6 — Implementation
Your time: 15 minutes per person
Deploy your ContextPrompt across AI tools. Onboard your team.
The biggest investment isn’t money or time. It’s trusting that the knowledge in your people’s heads is worth capturing. Everything else, we handle.
What you actually get: a Living Book
A Living Book isn’t a report. It isn’t a policy document. It’s your organization’s knowledge — written as narrative, structured for AI, readable by humans.
It captures what your best people know but have never written down: why they handle situation X differently from situation Y. What “good judgment” actually looks like in your context. The unwritten rules that make your organization work.
When loaded as AI context, a Living Book changes how AI responds. Not through rules or restrictions — through understanding. The AI starts to recognize situations the way your experienced staff does.
One document. Multiple functions. Living value.
How we capture what can’t be Googled
We don’t use workshops with Post-it notes. We don’t send questionnaires.
Our methodology is built on Cognitive Task Analysis — a research-grade approach developed by Gary Klein, Beth Crandall, and their colleagues to capture how experts make decisions under uncertainty. Originally designed for pilots, surgeons, and firefighters. Now applied to organizations navigating AI.
The method works because it doesn’t ask people what they think they do. It asks them to walk through specific decisions — step by step, in context — so the knowledge that lives between the lines becomes visible.
Validated across industries
Our methodology has been applied across logistics, education, public sector, and regulated industries. The Living Book format works in every domain because it captures how your people think — not just what they do.
Same structure. Your reality. Every time.
It’s like having a colleague who actually listened for three weeks — and remembers everything.
Description from a Living Book pilot project
What it costs to wait
Every month without AI alignment adds risk and waste.
Your AI gives inconsistent answers — eroding trust one interaction at a time. Knowledge walks out the door with every departure. EU AI Act enforcement is live, with the most significant compliance deadline — high-risk AI systems — arriving August 2026.
The numbers are clear: Fortune 500 companies lose an estimated $31.5 billion annually from poor knowledge sharing. Retroactive compliance costs 3–5x more than proactive. And a single AI compliance incident can cost more than your entire AI infrastructure.
A ContextPrompt engagement — €35,000 — captures, codifies, and deploys your organizational judgment in 6 weeks. A single compliance incident costs ten times that. A single key person leaving with 20 years of undocumented knowledge — that cost is incalculable.
The question isn’t whether you can afford to do this. It’s whether you can afford not to.
The math behind the investment
Consider a mid-sized organization: 200 employees, 100 active AI users.
The cost of the problem
Information search: Your people spend an estimated 20% of their time looking for information. At average knowledge worker rates, that’s over €1M per year in lost productivity.
Knowledge loss: With average turnover, the cost of lost tacit knowledge — the decisions, judgment calls, and institutional memory that walk out the door — runs into hundreds of thousands annually.
Compliance risk: EU AI Act penalties for non-compliance reach up to €35M or 7% of global revenue. Even a modest probability creates significant expected cost.
The investment
ContextPrompt: €35,000. Delivered in 6 weeks.
Even a 10% improvement in AI consistency, knowledge retention, and compliance readiness delivers ROI within the first year. Most organizations report impact from day one.
Start with a conversation
30 minutes. No pitch deck. We listen to where you are with AI, what’s working, what isn’t — and whether ValuesFilters is the right fit.
If it’s not, we’ll tell you. If it is, you’ll know exactly what happens next.