From “we should do AI” to “here is exactly what to build”
In one week we work alongside your leadership, operations and technology teams to answer four questions: where AI can create real value in your business, whether you are ready to build it, what to build first, and what to start in the next 90 days. You get a current-state assessment, an opportunity map scored on impact and feasibility, a 12-month roadmap and a 90-day plan, with an executive readout on Day 5.
- Engagement length
- One week, after a week of async preparation
- Executive readout
- Day 5
- Stakeholders interviewed
- 6 to 10 across leadership, ops, product, engineering and data
- What you leave with
- Opportunity map, 12-month roadmap, 90-day plan
- Start after signing
- Preparation can begin within two weeks
The moment you are in
Your board is asking about AI. Competitors are announcing things. Your team may have tried a pilot or two that did not go anywhere. And the honest answer to “where should we start?” still feels murky.
You are not alone. Most mid-market companies in 2026 are in this exact position. The market is loud, the vendors are eager, and every consultant will happily sell you a strategy deck. What is actually hard is figuring out which two or three AI initiatives will move your business, and what it will realistically take to ship them.
That is what this engagement is built for.
How this differs from the fixed-fee Week 1 audit
Two things on this site are called an audit and they are not the same size, so it is worth being blunt about which one you want. If you already know roughly what you want built, book the Week 1 audit and ignore the rest of this page. If your board is asking about AI and nobody can yet name the two or three initiatives worth doing, this is the piece of work that answers that. Either way, the first conversation is the same call.
Two calls and a one-pager, on one problem
Fixed fee, fixed scope, two calls. We take a single use case apart with the people who own it and write down what good means, where the risk sits, which decisions have to reach a human and what it would take to build. You keep the one-pager whether or not you continue. It is the opening move of a build, and it assumes the build is roughly known.
A week across the business, on the whole portfolio
One week of engagement with your leadership, operations and technology teams, preceded by a week of async preparation, with an executive readout on Day 5. Six to ten stakeholders interviewed, your data and tech landscape assessed, five to ten opportunities scored against each other, and a 12-month roadmap sequenced. This is for when the question is which AI work to do at all, and in what order.
The four questions the week answers
Not a strategy deck. Four answers, each one specific enough to act on or to argue with.
- Where
Where can AI create real value here
Not generic use cases. Specific opportunities in your business, scored by impact, feasibility and cost to build, in your context and against each other.
- Ready
Are you ready to build them
An honest look at your data, your tech stack and your team's readiness. What is in place, what is missing, and what has to happen first. Sometimes the answer is that a data problem has to be fixed before any of the AI is worth starting.
- First
What should you build first
A prioritised 12-month roadmap. Sequenced, realistic, and yours to own whether we build any of it or not.
- Now
How do you get moving
A concrete 90-day plan with one or two pilots we recommend you start now, scoped, with success criteria written down before anybody starts.
What you get
Six deliverables, all written, all handed over on Day 5.
Discovery sessions with 6 to 10 stakeholders
Across leadership, operations, product, engineering and data. We listen more than we talk.
Current-state assessment
Where your data lives, what your tech landscape looks like, and how ready your teams actually are for AI adoption rather than how ready the org chart suggests.
AI opportunity map
Five to ten prioritised use cases, each scored on business impact, technical feasibility and cost to build, so they can be compared rather than championed.
12-month roadmap
Sequencing, dependencies, resourcing, and rough cost ranges for the initiatives themselves so the plan can be budgeted instead of admired.
90-day quick-win plan
One or two pilots ready to start immediately, with scope and success criteria defined.
Executive readout
A clean 90-minute debrief for your leadership team, with everything written handed over at the end of it.
Before Day 1: one week of async preparation
An audit only works if the week itself is spent generating insight rather than gathering context. We move the context-gathering into a preparation phase before Day 1, so your team's time is spent on what only your team can provide.
Kickoff alignment call
Sixty minutes with your executive sponsor to confirm scope, stakeholders and success criteria.
Pre-read pack shared with us
Business overview, current tech stack summary, org chart of the AI-relevant teams, any prior AI initiatives including what did not work, and the problems top of mind for leadership. Standard NDA in place first.
Stakeholder scheduling
We work with your point of contact to confirm 6 to 10 interview slots across the engagement week. Executives, ops leads, product, engineering, data.
Discovery questionnaire
A short async form for each stakeholder to complete before their interview. Interviews then go deep instead of staying at introductions.
Systems and data access
Read-only access to dashboards, data catalogues and relevant documentation, set up ahead of Day 1 so no engagement time is lost on plumbing.
The week itself, day by day
Five working days, with the readout fixed at the end of them rather than promised for later.
- Day 1
Kickoff and first interviews
Kickoff meeting with your sponsor and core team, followed by two or three stakeholder interviews.
- Days 2 and 3
Deep discovery
The remaining stakeholder interviews, the data and tech stack review, and workflow observation where watching the work teaches more than asking about it.
- Day 4
Synthesis and assessment
We synthesise the findings, score the opportunities and draft the roadmap. Interim check-in with your sponsor to validate direction, so nothing in the readout is a surprise to the person who commissioned it.
- Day 5
Executive readout and delivery
A 90-minute readout with your leadership team. Every written deliverable handed over on the day.
- After
Follow-up, and an optional handoff to build
One 30-minute call seven days after the readout to answer questions and support internal alignment. If you choose to move to implementation, the same team continues without re-scoping or re-discovery. If you do not, the roadmap is still yours.
What you need to bring
Three things from your side. Everything else is on us.
- 01
An executive sponsor who owns the outcome
A CEO, CTO or COO in most cases. Somebody whose own plan changes depending on what the roadmap says.
- 02
A point of contact who can open doors
Somebody who can schedule stakeholder time and get access to the relevant documents and systems without a fortnight of chasing.
- 03
Honest access to your teams
If people give us safe, filtered answers, the roadmap will not reflect reality. This is the one input we cannot supply ourselves, and the one that decides whether the week is worth anything.
How we are different
Most AI consulting is done by people who have never shipped AI. That is the industry's quiet secret. They read the papers, interview the vendors, and hand you a strategy deck. Then a different team, often junior or offshore, is left to figure out how to actually build it. That handoff is where most AI initiatives die.
We are the team that ships. When we assess your business, we bring the perspective of engineers who deploy AI into production every day. Our recommendations are shaped by what is actually possible in your stack, on your timeline, with the people you have. And if you decide to build with us afterwards, you get the same people from audit to production. No handoff. No re-scoping. No lost context.
Who this is for
And, by omission, who it is not for. If you are a twenty-person company with one clear use case, the Week 1 audit is the better fit and costs you far less time.
Mid-market companies exploring AI seriously
Roughly 100 to 5000 employees, where there are enough moving parts that the opportunities have to be compared rather than simply listed.
Leadership teams who need clarity before committing engineering budget
The decision is not whether to spend, it is where, and nobody wants to defend the answer without evidence behind it.
Companies whose first pilots fizzled
One or two attempts that went nowhere, and a wish to reset with a plan rather than try a third thing at random.
Product companies who know AI belongs in the roadmap
The intent is settled and the specifics are not, which is a harder position to get out of than it looks.
Family businesses and traditional enterprises with a modernisation mandate
Where the mandate arrived before the plan did, and the plan now has to be defensible to people who are sceptical.
The timeline, in four numbers
We keep the engagement fast and focused because slow assessments do not ship. A four-month AI strategy delivered in Q3 helps nobody.
About Woodfrog
Who is actually in the room for the week.
Anthropic Build Partner
We are an Anthropic Build Partner and joined the Claude Partner Network at launch. We ship production AI on Claude, which is why what the roadmap recommends is bounded by what we know deploys rather than by what reads well.
18 people, and a pod of three per engagement
Every engagement runs with a pod of three experienced practitioners, no bench and nobody learning on your clock. The people who assess your business are the people available to build what the roadmap says to build.
50+ projects delivered, 20+ clients
Across manufacturing, fintech, retail and SaaS since 2023, which is where the pattern library for scoring your opportunities comes from. Opportunities get scored against work we have actually delivered rather than against a market report.
Pune, India, founded in 2023
We build production analytics, AI and data platforms for enterprises and product companies, and we deliver from Pune across India, APAC and the US. Our median time to a first artefact running in production is six weeks, which is the pace the roadmap's sequencing assumes rather than a pace it hopes for.
The questions sponsors ask before signing
Including the ones where the honest answer is no.
How is this different from the Week 1 audit you sell everywhere else?
Scope, not depth of care. The Week 1 audit is two calls and a written one-pager on one use case you have already identified, at a fixed fee, and you keep it either way. This is a full week with 6 to 10 people across your business, preceded by a week of async preparation, and it compares opportunities you have not identified yet. Most companies want the Week 1 audit. You want this one when nobody can yet name which AI initiatives are worth doing.
How much of our people's time does this actually take?
About an hour each from 6 to 10 stakeholders, plus a short questionnaire before their interview. Your executive sponsor gives a 60-minute call up front, an interim check-in on Day 4 and the 90-minute readout on Day 5. Your point of contact carries the scheduling and access work in the preparation week. Nobody else has to clear a week.
Do we have to build with you afterwards?
No. The roadmap is yours to own, including if you take it to another firm or build it in-house. The optional handoff exists because re-discovery is waste, not because the assessment is a sales exercise with a deliverable attached.
What if the honest answer is that we are not ready for AI yet?
Then that is the finding, and it is written down as plainly as any other. Readiness is one of the four questions the week answers, and a roadmap that opens with fixing a data foundation is a real answer. It is also a cheaper answer than discovering the same thing nine months into a build.
We tried a pilot that went nowhere. Why would this be different?
Failed pilots are one of the most useful inputs we ask for in the pre-read pack, including what specifically did not work. Most of them fail for a reason that repeats: no owner, no success criteria written before the start, or a use case chosen because it demonstrated well rather than because it mattered. Naming which of those happened is part of the assessment.
Will you look at our data, or only interview people?
Both. Read-only access to dashboards, data catalogues and documentation is set up before Day 1, and Days 2 and 3 include the data and tech stack review alongside the interviews. An assessment built only on what people say about their data is a survey, not an audit.
Why one week rather than a proper three-month strategy engagement?
Because slow assessments do not ship. A four-month AI strategy delivered in Q3 helps nobody, and the marginal insight after week one comes mostly from repetition. The preparation phase is what makes one week enough: by Day 1 we already have the context a longer engagement spends its first month collecting.
Who from your side is in the room?
A pod of three experienced practitioners, no bench, and the same people who would build the work if you continue. We are an Anthropic Build Partner and joined the Claude Partner Network at launch, so model and architecture recommendations come from engineers who deploy on Claude rather than from a vendor briefing.
Can you run this if our leadership team is spread across sites?
Yes. Interviews are scheduled around your people rather than around a room, and we deliver from Pune across India, APAC and the US. The two fixed points are the kickoff and the Day 5 readout, which we want live with your leadership team together, in person where that is practical.
The rest of the practice
These are genuinely different jobs with different ways of failing. Most engagements start in one of them.
- Data engineeringJobs that finish, migrations that reconcile, and storage that stops paying for cold data.
- Data integration and governanceOne set of records the finance team and the operations team both accept.
- Data platforms and modernisationOff the platform you outgrew, without a twelve-month freeze on new reporting.
- Apache SupersetSuperset built and run by people who commit to the project, including embedding and Kubernetes.
- Command centresThe one screen an operations floor runs the day from, not another dashboard.
- AI evaluationA graded eval set, because an AI system cannot simply pass or fail a test suite.
- AI governanceGovernance you can defend in a board meeting and audit on demand.
- AI agentsOne reasoning agent on Claude, on your live data, with evidence and a kill switch.
- Data agentsAgents that watch the data, flag what moved, and explain what is driving it.
- AI use case, guaranteedOne production agent in eight weeks, or you do not pay for the build.
- Applications and automationThe system your team works in all day, built or replaced in slices.
- Application modernisationThe system nobody wants to touch, replaced a slice at a time rather than rewritten.
- System integrationSystems that stop disagreeing about the same customer, order and item.
- Process digitisationThe process that still runs on paper, WhatsApp and one shared spreadsheet.
Start with the call, then decide which of the two you actually need
Thirty minutes. We listen to where you are, tell you plainly whether the Week 1 audit or the full audit and roadmap fits your situation, and answer questions. No obligation. If you already know you want the full week, the preparation phase can start within two weeks of signing. Email hello@woodfrog.tech. We work under NDA and to a fixed scope, from Pune, across India, APAC and the US.