Stop Token :: machine-readable index
Plain-text mirror of the site for agents, crawlers, and anyone who would rather read it raw. Same content, no layout to infer meaning from.
- /Home
- /#auditAI Readiness audit
- /#planningAI Transformation Planning
- /#common-senseAI of Common Sense
- /#adoptionAI Adoption programmes
- /#governanceBI / Data / AI Governance Strategy
- /enoughThe ENOUGH framework
- /readinessReadiness check
- /stackAI stack chooser
- /m/stackStack, machine view
## About
NameStop Token
PositioningCommon sense AI agency
PeopleMax Ivanov, Alex Barakov
ModelTwo senior practitioners. Whoever you meet on the first call does the work.
Most AI projects fail from too much, not too little. We find the smallest system that moves a number you already report on, build it, and stop. Two of the 5 services exist to stop a client spending.
## Services (5)
AI Readiness audit
PromiseWhere AI pays off here, and what not to buy.
ForWhoever holds the AI budget — CTO, head of data, the AI or BI office
Duration4–6 weeks
OutcomeA ranked list with the numbers under each candidate, a written case for and against, and the gaps to close before anything starts.
Topics
- who uses which AI and analytics tools today, how often, for what
- workflows ranked by hours returned against cost to build and run
- data and access per candidate: what blocks it, for how long
- tooling: overlaps, licences nobody uses
- lock-in and the cost of leaving each proposed stack
- who would run it after it ships
- security, compliance and the approvals you will need
- how much your team can absorb at once
AI Transformation Planning
PromiseTwelve months of what to build, and what to refuse.
ForThe executive team and the CDO or CTO office
Duration4–6 weeks
OutcomeA twelve-month roadmap with people and infrastructure per bet, the metrics that will show whether it worked, and the list of what not to build.
Topics
- operating model: what stays central, what moves to the teams
- the portfolio: sequence, dependencies, what each needs to start
- economics: hours returned, cost to run, ROI with named assumptions
- measurement in the reporting you already run, not in a slide
- roles: who you hire, who you train, who you stop hiring
- platforms: build, buy or wait, with the reasoning written down
- rules and risk limits for AI agents acting in your systems
- what you will not do, and why
AI of Common Sense
PromiseThe smallest set of tools and practice that lets your team go on without us.
ForThe AI or data office, BI teams, and the engineers who will own it
Duration5–8 weeks
OutcomeA working set-up in your infrastructure, tests that prove it works, all of it in your ownership, and a runbook: monthly cost, failure points, what to watch. Then we leave.
Topics
- skills: reusable instructions for AI agents, written by your team
- connectors from agents to your systems (MCP), access limited and logged
- one gateway for every AI model: budgets, fallbacks, one bill
- the runtime for agents: how they run, retry and get watched
- context: what an agent may know and where that lives
- evaluation: test cases and a score that shows whether a change helped
- duty bots: routine on-call and triage handed to agents
- data assistants that answer from agreed metric definitions, not raw tables
- which jobs are worth an agent, and which are not
AI Adoption programmes
PromiseTraining that ends in shipped agents, not certificates.
ForThe AI or data office, BI and analytics teams, and whoever bought the licences nobody uses
Duration6–8 weeks
OutcomeAgents and skills in production, built by your people; hours saved per participant; a written standard and facilitators on your side. Last accelerator: 60 people, 39 projects, 29 shipped, 8 reused.
Topics
- teams of two to four pick a real task from their week and build an agent or a skill for it
- weekly sessions with mentors, a demo every week, a pitch to a jury at the end
- the rubric: shipped or not, data safety, measured effect, reusable by others
- the result: agents in production and a catalogue of who reuses them
- one department, the same modules, everyone on their own tasks
- a baseline before, hours saved per person after
- a standard the participants write: how AI work is checked before it counts
- facilitators trained inside, so the next run is yours
BI / Data / AI Governance Strategy
PromiseA strategy session that ends in a governance roadmap your own people agreed to.
ForBI, data and ML leadership, the CDO office, and the business leaders who depend on the numbers
Duration4 weeks
OutcomeA governance roadmap with owners, dates and a detailed first quarter; the roles and rules agreed in the session; a maturity reading to measure the next one against; and the session's materials.
Topics
- preparation: up to ten interviews, usage statistics, the documents that exist
- a maturity reading across data, analytics and AI
- the session: half-day workshops with the leadership and the people who do the work
- ownership: data owners, metric curators, AI champions — names, not committees
- the rules: data quality, metric definitions, model and agent use, access, compliance
- demand against supply: what the business asks for, what the teams can deliver
- the sequence: what first, what later, what not at all
- how decisions reach the people they affect
## How we work — ENOUGH
- E — Examine. What is already true: readiness, the jobs to be done, what has been bought
- N — Narrow. What to build, in what order, and what will not be built
- O — Outline. The smallest architecture that carries the chosen work
- U — Use. One or two things in production, used in real weeks
- G — Gauge. What it returned and what it costs to run
- H — Hand over. The roadmap the team runs without us
## Readiness check — 4 layers, 16 checks
A chain, not a score. Each layer is measured against a target set by six facts about the company, and the chain is only as good as the first layer that does not hold.
Foundation — the tables the answers come from
- Every metric in your main report has a named owner — a person, not a team. (10 of 10; hours)
- The tables behind them have described columns. (a third or more; days)
- Each of those tables has an automated check that runs. (one per table; days)
- Data arrives on a stated time, and no silent delay broke a report last month. (SLA stated, zero silent breaks; weeks)
Semantic — one meaning per number
- Each core metric has one written definition, and two teams get the same number from it. (top 3 metrics match; days)
- The definitions live in one place a machine can read. (one source, not one per dashboard; weeks)
- A glossary of the domain's terms exists and is used. (15+ terms, edited this quarter; days)
- Key metrics have written reading rules: what is normal, what is a warning, which comparisons are allowed. (top 5 metrics; days)
Knowledge — what an agent would need to be given
- Recurring questions have written answers a machine could read, and the set is current. (30+, edited this quarter; weeks)
- Test cases exist — a question and its expected answer — and they have been run against what you have now. (10+, scored this quarter; days)
- Worked examples exist: a real question, the query that settled it, and the source the answer can be checked against. (5+, each traceable; days)
- No change to a prompt, a model or a retrieval setting ships without being scored against those cases. (a written rule, and the last change scored; days)
Operation — people and gates
- Three people are named: data owner, metrics curator, AI champion. (three names, three jobs; hours)
- Somebody has hours in the week to judge machine output, as their job. (hours in a calendar or a role; a hire)
- No machine-produced number reaches a decision without a named person signing it off. (a written rule; days)
- AI spend is visible in one place with an owner, and you can state what one accepted answer costs. (one bill, one name, a cost per accepted answer; days)
Levels0 Not yet · 1 Foundation ready · 2 Semantic ready · 3 Knowledge ready · 4 Operation ready
Human view/readiness
## Stack map
Tools217
Layers15
GroupsModel, Context, Agent, Access, Run & Cost, Control
Open source109
Full listing/m/stack
## Contact
Emailinfo@stop-token.com
WhatsAppopen a chat
Telegramopen a chat
AddressDelft, Netherlands
Privacy/privacy
2026 Data Nature · KVK 90218426 · Common sense AI agency