A Power BI consultant for models, DAX and reports people trust
When a Power BI report is slow, shows a different number from the one in finance's spreadsheet, or shows data to people who should not see it, the problem is rarely the visual. It is the model underneath, the measures and the security. As your Power BI consultant, we fix those: a clear semantic model, DAX measures defined once, row-level security that matches your organisation, reliable refreshes through gateways, and the right licence and capacity. Everything stays in your own Microsoft tenant.
- No minimum contract: book a few hours or a full project
- Scope agreed in writing before work starts
- Independent advice: no reseller commissions
When to bring in a Power BI consultant
- Reports take many seconds to load, or refreshes fail at night and nobody notices until the morning meeting.
- The same figure, such as revenue or margin, differs between two reports, because each report has its own version of the calculation.
- Every new report starts with a new import from the same source, and nobody knows which dataset is the right one.
- People see data they should not, or the security is managed by copying reports per team.
- You are on a Premium P SKU, or thinking about Microsoft Fabric, and need a sizing and migration plan.
- The person who built the reports has left, and nobody dares to change them.
What we do
- Semantic model design. A star schema with fact and dimension tables, clear relationships, a proper date table and only the columns reports need.
- DAX measures. Business definitions such as revenue, margin, churn or utilisation written once in the model, documented and reused by every report.
- Power Query and data preparation. Transformations moved as close to the source as possible, with query folding and incremental refresh for large tables.
- Row-level security. Roles and rules that match your organisation, tested with real users, and workspace permissions that do not undo them.
- Gateways and refresh. On-premises data gateways set up, monitored and documented, with alerts when a refresh fails.
- Workspaces, apps and governance. Development, test and production workspaces, deployment pipelines or Git integration, and apps for distribution instead of shared links.
- Performance tuning. Slow visuals traced to the measure or the model, storage modes chosen per table, and aggregations where volumes need them.
- Migrations. From Excel reporting or another BI tool to Power BI, and from Power BI Premium P SKUs to Fabric capacity.
Import, DirectQuery or Direct Lake
The storage mode decides speed, freshness and cost, and it is chosen per table. Import copies the data into the model and is fastest to query, at the cost of scheduled refreshes. DirectQuery leaves the data in the source and queries it live, which keeps it fresh but moves the load to the source system. Microsoft describes Direct Lake as a storage mode in Microsoft Fabric that loads data from Delta tables in OneLake into memory, with query performance comparable to Import mode but without refresh cycles that copy the entire data volume. (Read on 11 November 2026.) Direct Lake makes sense when your data already lives in a Fabric lakehouse or warehouse; for a company whose data sits in an ERP and a few cloud apps, Import with incremental refresh is usually the simpler, sturdier choice.
Security that matches your organisation
Microsoft's documentation describes row-level security as filters, defined within roles, that restrict data access for specific users of a semantic model, and notes that the roles apply to members of a workspace in the Viewer role. (Read on 11 November 2026.) In practice, that means two things must be designed together: the RLS roles in the model, and who gets which role in each workspace. A common mistake is to give report readers Member or Contributor access for convenience, which bypasses row-level security entirely. We set up roles from your organisation data, such as region, cost centre or customer, test them as real users, and document who sees what.
Gateways and data sources
Data in an on-premises database or file share reaches Power BI through a gateway. Microsoft describes the on-premises data gateway as locally installed Windows software that connects the Power BI service to data sources in your network using outbound connections, so no inbound ports need to be opened. (Read on 11 November 2026.) We install gateways in a cluster where refreshes are business-critical, run them under service accounts in your name, keep credentials out of personal accounts and set alerts for failed refreshes, so a night's refresh failure is a message in the morning, not a wrong number in the board pack.
Power BI Premium, Fabric and licences
Licensing is where Power BI costs are decided, and it has changed. Microsoft's migration FAQ states that each Power BI Premium P SKU subscription retires at the end of its current agreement term, that new P SKUs are no longer sold, and that migration is not automatic: you buy a Fabric F SKU in Azure and reassign each workspace. If a P SKU ends with no F SKU in place, content enters a 30-day grace period, is throttled from day 31, and from day 91 all operations are rejected. The same FAQ explains that on F64 and larger capacities, users with a free licence in the Viewer role can view content, while on F2 to F32 every viewer needs a Power BI Pro or Premium Per User licence. (Read on 11 November 2026.)
The right setup depends on how many people build reports, how many only read them, and how large the data is. We count those from your tenant's usage data, compare the options in licence terms, and recommend the smallest setup that works. We do not resell licences; you buy them in your company's name. Our guide to dashboard development cost shows how build, licences and upkeep add up over five years.
Power BI or a custom dashboard?
Power BI is the right tool when the audience is your own staff, the questions change often and your analysts want to build their own reports. A custom dashboard is better when the audience is your customers inside your product, when the screen must work like an app, or when per-user licences for a large audience do not make sense. We build both; for the second, see dashboard development. If the question is what to measure in the first place, start with data analytics consulting.
How an engagement runs
- Review. We read your tenant: workspaces, models, refresh history, gateway health, licences and the reports people actually open. You get a short written finding list.
- Plan in writing. What to fix first, the target model and the hours, agreed before work starts.
- Model first. The semantic model and measures are rebuilt or cleaned, with the old reports pointed at the new model one by one.
- Security and refresh. Roles, workspace access, gateways and alerts, tested with real users.
- Hand-over. Documentation of the model and measures, and working sessions with your team on your own data.
You own it
Everything lives in your company's Microsoft tenant: workspaces, semantic models, reports, gateways and licences. Project files are kept in Power BI project format in your own repository, so changes are reviewed and versioned like code. We work through accounts you create and remove when the work ends, and nothing is locked to us. See prices.
Questions we get
Can you fix our existing Power BI reports rather than start again?
Usually, yes. Most slow or inconsistent reports share a few causes: a model without a clear star schema, measures duplicated in every report, and refreshes that load more than they need. We fix the model first, then point the existing reports at it, so users keep the pages they know.
Do we need Microsoft Fabric?
Not necessarily. Power BI Pro or Premium Per User covers many companies. Fabric capacity is worth it when many people only view reports, when data volumes outgrow Import mode, or when you want lakehouse storage and Direct Lake. We look at your users and data and recommend the smallest setup that works.
We are on Power BI Premium P SKU. What do we have to do?
Microsoft no longer sells P SKUs and each subscription retires at the end of its current agreement term. You buy a Fabric F SKU in Azure and reassign your workspaces to it before the P SKU ends. We plan the size, the order of workspaces and a test window, so nothing falls into the grace period.
Can Power BI show each user only their own data?
Yes, with row-level security in the semantic model. Note that it restricts users in the Viewer role; workspace members with higher roles see all data. Getting the roles and the workspace permissions right together is part of the work.
Can you train our team?
Yes. Most engagements end with your analysts maintaining the model and building their own reports. We document the model, the measures and the refresh setup, and run short working sessions on your own data.
Who owns the reports and models?
Your company. Everything lives in your own Microsoft tenant and workspaces, and the project files sit in your repository. We work with accounts you create and remove when the work ends.
Get a written quote
Tell us in a few lines what you need. We reply within one working day with a written quote: scope, timeline and a fixed price or an hour estimate. Asking is free and commits you to nothing.