No price list. Two companies can ask for the same dashboard and need a week apart in effort, so a published number would either be wrong or would push us to cut corners to protect it. What you get instead is something more useful: a written scope, and a fixed quote you can hold us to.
If your budget is fixed and public, say so in the first email. We will tell you on the call whether the scope fits inside it, and if it does not, what part of it would.
You are not asked to sign anything on the call, and there is no discovery phase billed by the hour before you know what you are buying.
You describe the number or the process that is causing the problem. We tell you what usually causes it, which service line fits, and what a sensible first step would be. No deck, no commitment, and often the first step is smaller than you expected.
Within two working days you get a scope document: deliverables, the gates, what we need from your side, the assumptions we are working to, and what is explicitly out of scope. It is short enough to read in one sitting.
Priced against that scope. Systems, data quality, volume, tooling and how much of the work we can do remotely all move the number, and the scope says which of those we assumed. If any of them changes materially, we re-quote in writing before doing the work.
Larger builds are split across the Five Gates with a milestone against each, so you are never funding more than one gate ahead of delivery. Nothing is billed for a phase that fails its written exit criterion.
Published so you can estimate your own effort before you talk to us, and so the difference between two quotes makes sense.
One ERP is not the same job as an ERP, a CRM, a payroll system and four spreadsheets maintained by people who left.
If a source is unusable, cleaning it is its own piece of work and we will quote it separately rather than hide it in an estimate.
A client who already has a documented metric layer needs less of our time in Gate 01. That is usually the single biggest difference between two quotes for the same-looking project.
Monthly batch, nightly batch or streaming are three different builds. So is a model over 2,000 documents against one over two million.
A standard cloud deployment and an in-tenant deployment with your own keys and a security review are not the same amount of work.
If your team builds the pages and we build the layer, the number drops. We will tell you when that is the better split for you.
Which model fits is part of the scoping conversation, not a menu you have to choose from alone.
One price against one written scope and one end date. Most work runs this way.
Use it when: A defined deliverable: a metric layer, a warehouse migration, a dashboard set, an extraction pipeline.
A short, cheap first engagement that ends in a decision rather than a commitment.
Use it when: When the scope is not yet knowable: AI use cases, data quality, whether a build is worth doing at all.
A number of days per month, agreed in advance, used for whatever the roadmap needs that month.
Use it when: Definition changes, new pages, pipeline upkeep, and a named owner for the model.
Stated up front because these are the parts clients get surprised by elsewhere.
Fees exclude VAT and any travel. Engagements are delivered remotely from Dhaka with a standing overlap into European mornings; on-site presence is quoted separately if you want it.
Every service line below is scoped and priced client by client. The list is the shape of the work, and the durations are the honest range we see.
| Service line | Practice | Typical duration | Engagement |
|---|---|---|---|
| Metric audit | Metric foundation | 5 working days | Fixed-scope diagnostic |
| Semantic layer build | Metric foundation | 3 to 6 weeks | Fixed-price project in milestones |
| Metric ownership & governance | Metric foundation | 2 to 3 weeks | Fixed-scope workshop series |
| Warehouse & modelling | Data platform | 4 to 8 weeks | Fixed-price project in milestones |
| Integration & pipelines | Data platform | 3 to 8 weeks | Fixed-price or time and materials |
| Data contracts & quality | Data platform | 3 to 5 weeks | Retainer or fixed-scope project |
| Executive & operational dashboards | Reporting & enablement | 3 to 5 weeks | Fixed-price project |
| Planning & decision support | Reporting & enablement | 4 to 6 weeks | Fixed-price project |
| Team enablement | Reporting & enablement | 2 to 4 weeks | Fixed-scope programme |
| Two-week AI readiness sprint | AI · Readiness | 10 working days | Fixed-scope sprint |
| Governed natural-language BI | AI · Build | 3 to 6 weeks | Readiness sprint, then rollout |
| RAG over your documents and databases | AI · Build | 4 to 8 weeks | Fixed-scope build, then retainer |
| Agentic workflows for operations | AI · Build | 4 to 6 weeks to first production loop | Pilot to production, fixed value |
| Forecasting & anomaly detection | AI · Build | 4 weeks pilot | Fixed-scope pilot, then production |
| Document intelligence | AI · Build | 5 to 8 weeks | Fixed-scope build |
| EU AI Act & AI governance readiness | AI · Readiness | 3 to 5 weeks | Fixed-price assessment |
| MLOps & model monitoring | AI · Production | 4 to 6 weeks | Fixed-scope build, then retainer |
17 service lines. Scope document and fixed quote within two working days of the fit call. Replies within one working day.
The cheapest way to find out: start with the smallest thing that settles the argument, usually a metric audit. If the definitions are already sound, we will tell you that in the first week and you will not need the build.