Process

A short and honest description of what working with us looks like. What we'd ask of you, what we'd charge, how long things take, and when we'd say no.

  1. Discovery

    One to two weeks

    Start with a paragraph from you, then a call. By the end of the first conversation we should both know roughly whether we're the right people, what we'd build first, and what it would cost. If the answer is no, we'd say so on the call.

    If we're a fit, we'll write a short brief, half a page, plain English, describing the work, the deliverable, the timeline, and the price. You'll have it within three working days of the call.

  2. Build

    Four to twelve weeks for a Power BI rebuild  /  six to twelve months for a Fabric platform

    Most reporting and Power BI rebuilds run four to twelve weeks. A Fabric platform built from scratch is a different shape: six to twelve months, with the longer end accounting for stakeholder reviews, security sign-offs, and integration with the source systems you already have.

    We work in two-week sprints, demo at the end of each, and don't disappear between them. Daily Slack or Teams, weekly written progress note. You'll know what's happening without having to chase.

  3. Handover

    Built into the build, not bolted on at the end

    We don't believe in deliver-then-disappear. The handover starts in week one: documentation written as we go, your team brought into the design conversations, peer reviews you can sit in on.

    By the end you should be able to maintain the work, extend it, and break-and-fix without us. If you can't, we haven't finished.

  4. After

    A quiet first year

    We'll check in once a quarter for the first year. Free. If something has broken, or if you need a small change we can do quickly, we'll quote it. If it turns into a bigger second engagement, that starts at Discovery again.

Power BI

What a Power BI rebuild covers

Most rebuilds run four to twelve weeks. Smaller refits at the lower end, full reporting redesigns at the upper end. Inside that:

  • Optimised DAX measures: rewritten for speed, readability, and maintenance
  • Reports, dashboards, paginated reports, and metrics, built to align with Gestalt principles of design
  • Accessibility built in: screen-reader friendly, keyboard navigable, contrast-checked
  • Star-shaped schemas for read-optimised data models
  • Basic data quality and hygiene at source
  • Optimised Power Query steps
  • Review and rationalisation of data sources
  • Training delivered to your in-house analysts
  • Broken Power BI reports fixed, including the ones nobody wants to touch
  • Administration of sharing, security, organisational and workspace apps
  • Licence management
  • Row-level, object-level, and column-level security (RLS, OLS, CLS)
Fabric

What a Fabric implementation involves

A Fabric platform from scratch usually runs six to twelve months. The longer end accounts for the things that always take longer than expected: stakeholder reviews, security sign-offs, integration with the source systems already in place. Inside that:

  • Slowly Changing Dimension Type 2 (SCD2) handling
  • OneLake security and access policies
  • Lakehouses and warehouses, structured to fit the work
  • Medallion architecture: bronze, silver, gold layering
  • Direct Lake semantic models for sub-second reporting
  • Advanced, efficient pipelines with approval activities, web hooks, OAuth2, and notifications
  • A data quality engine that catches issues before they reach the reports
  • Advanced data extraction, including incremental refreshes and page-fetching from APIs
  • API, Azure, and cloud service integrations
  • Sensitive and confidential data protection mechanisms
  • Advanced data governance
  • Training, upskilling, and peer reviews
  • Migration from legacy or outdated ETL systems
  • Power BI reports built on top of the platform
  • Real-time event data processing and analysis
  • Adoption work and data culture support
  • International rollout with culture-aware standards
  • Accessibility throughout
  • Dataflow Gen 2 and notebooks where each earns its place
  • Optimised processing for big data using advanced Apache Spark techniques
  • Capacity sizing and cost optimisation
  • Git integration and deployment pipelines
  • Monitoring and alerting via Data Activator
On price

What we charge

We quote either a day rate or a fixed price, depending on which is fairer for the work. Fixed price when the scope is clear. Day rate when it isn't.

£950 to £1,350 Day rate, standard work

We don't bill for the discovery call. The first three days of any engagement are at our standard rate. If it's not working out by day three, we'd rather end there cleanly than pad the time.

When it shifts

Estimates aren't promises

Sometimes the work takes longer than we thought, or scope changes, or something we hadn't accounted for surfaces. When that happens we'd rather tell you early than late: candidly, with the numbers, and with options.

Most of the time the right answer is a quick parley to reset scope, deadline, or budget so the work still lands well, on terms that are fair to both sides. We'd rather have an awkward conversation in week six than a worse one in week ten.

On fit

When we say no

It's better to say no early than to disappoint later. We'd say no if any of these are true:

  • The work would only need a junior contractor. We'd help you hire one.
  • The deadline can't be met by anyone, and we don't think we should be the ones who pretend otherwise.
  • The brief is "make our data work" with no shape behind it. We'd ask you to come back when there's a question we can answer.
  • The work is a vendor pitch dressed up as a problem. We don't do those.
On the work

What "still useful after we've left" means

It means the team you trained still owns the work. The pipelines still run. The reports still trust their numbers. The next person to touch the codebase doesn't curse at it.

It also means we'd rather walk away from a sale that doesn't work than win one that doesn't last.

Tell us what you're working on.

A paragraph is plenty. We usually reply within two working days, with our take on whether we're the right people, where we'd start, and roughly what it'd cost. If it's not a fit, we'll say so, and point you somewhere better.

Write to zak@tahleelconsultancy.com