Microsoft services

Microsoft Fabric: the unified data platform, implemented properly

Microsoft Fabric brings data engineering, warehousing, real-time intelligence, data science, and Power BI into one SaaS platform on OneLake. Veratas designs and builds enterprise Fabric estates, and migrates analytics off Azure Synapse and Power BI Premium, with capacity and governance designed in so the platform performs without over-spend.

Talk to our team
The platform

What makes Fabric different

Fabric is not a rebadge of Synapse. The architecture changes how a data estate is built.

Fabric is a software-as-a-service analytics platform built on OneLake, a single tenant-wide data lake. Every workload, the Lakehouse, the Warehouse, Power BI semantic models, KQL databases, writes to the same OneLake storage in open Delta Parquet format. That means data is stored once and read by many engines, instead of being copied between a lake, a warehouse, and an analysis cube. Shortcuts let a workspace reference data held elsewhere, including Azure Data Lake Storage and Amazon S3, without physically moving it.

The other shift is commercial and operational. Fabric is bought as capacity, an F-SKU measured in capacity units, and that single capacity is shared across all workloads. There is no separate cluster to size for Spark, no dedicated SQL pool to pause. This makes Fabric simpler to run, but it also means capacity sizing and consumption monitoring are central engineering tasks, not afterthoughts. A Fabric estate designed without capacity discipline either throttles under load or quietly burns budget.

Fixed-fee implementation

Microsoft Fabric, foundation set, from $10,000

A fixed-scope Microsoft Fabric foundation, from workspace to your first report. You see exactly what the price covers, and a fair quote for anything more.

Fixed fee from $10,000   |   Live in 4 to 6 weeks   |   Price shown up front

Get your fixed quote
What $10,000 covers

A working Microsoft Fabric foundation

A governed Fabric workspace with a lakehouse, pipelines, and a first report.

  • A Microsoft Fabric workspace, set up and governed
  • One lakehouse with a medallion (bronze, silver, gold) structure
  • Two data pipelines from your sources
  • A semantic model over the gold layer
  • One Power BI report on the model
  • Basic workspace governance and access

Microsoft Fabric capacity is billed separately by Microsoft or through CSP. The $10,000 covers the implementation services.

No surprises

What is in the $10,000, and what we quote separately

Anything beyond the standard package is optional, and always quoted before you commit.

In your $10,000Beyond the package, quoted at a fair rate
One lakehouse and two pipelinesMultiple workspaces and source systems
One semantic model and reportReal-Time Intelligence and event streams
Medallion structureA full enterprise warehouse
Basic governanceMachine learning and data science
First-report enablementGovernance at scale and a centre of excellence

You pay $10,000 for the foundation. Everything else is optional, scoped and quoted transparently at a reasonable rate, and always shown before you decide.

What we deliver

End-to-end Microsoft Fabric delivery

Veratas covers the full Fabric platform, from OneLake foundations through medallion modelling to Direct Lake reporting.

OneLake and workspace architecture

OneLake foundation design, the domain and workspace topology that matches how your organisation is structured, capacity allocation, and a shortcut strategy so data is referenced rather than copied across teams.

Lakehouse and warehouse modelling

The right store for each workload: Lakehouse for data engineering and Spark, Warehouse for T-SQL and multi-table transactions. Both modelled as a medallion architecture with bronze, silver, and gold layers.

Data engineering and pipelines

Data Factory pipelines, Spark notebooks, and dataflows for ingestion and transformation, with on-premises connectivity through the data gateway and full run history, retry, and alerting.

Direct Lake and Power BI

Power BI semantic models in Direct Lake mode, reading Delta tables straight from OneLake with no import refresh and no DirectQuery latency, the pattern that scales to thousands of concurrent users.

Real-Time Intelligence

Event and streaming workloads through Eventstream, Eventhouse, and KQL databases, with real-time dashboards and Reflex alerts where decisions cannot wait for a batch cycle.

Synapse and Premium migration

Structured migration from Azure Synapse Analytics and Power BI Premium onto Fabric, consolidating engineering, warehousing, and BI on one platform and one bill.

Architecture

Lakehouse or warehouse, and why the medallion layout matters

The two stores look similar from a distance. Choosing well, and layering well, keeps the estate maintainable.

Both the Fabric Lakehouse and the Fabric Warehouse store data as Delta tables in OneLake and both are queryable in T-SQL, but they suit different work. The Lakehouse is the home for data engineering: Spark notebooks, unstructured and semi-structured data, and large-scale transformation. The Warehouse is a full T-SQL engine with multi-table transactions and the developer experience a SQL team expects. Most enterprise estates use both, the Lakehouse for ingestion and heavy transformation and the Warehouse for the curated serving layer, and the right split is an architecture decision we make with you, not a default.

Across both, we structure data as a medallion architecture. The bronze layer holds raw ingested data exactly as it arrived, for replay and audit. The silver layer is cleansed, conformed, and deduplicated. The gold layer holds business-ready dimensional models that semantic models and reports consume. This layering keeps lineage clear, makes pipelines restartable, and means a change in source logic is contained rather than rippling straight into reports. It is the difference between a Fabric estate that stays understandable at year three and one that becomes a tangle.

How we deliver

A proven Fabric implementation path

Fabric estates are built through five phases, with capacity, governance, and cost control designed in from the first.

01

Assess

Current data estate review, source and report inventory, capacity sizing modelled against real workload, and a target Fabric architecture. We output a fixed-price design and, where relevant, a migration plan off Synapse or Premium.

02

Foundation

OneLake, domain, workspace, and capacity setup; a Microsoft Purview governance baseline for classification and lineage; the security and access model; and CI/CD through Fabric deployment pipelines linked to source control.

03

Build

Lakehouse and Warehouse modelling, Data Factory ingestion pipelines, Spark transformation, and medallion-layer development, delivered in agile sprints with working data demonstrated each cycle.

04

Report

Power BI semantic models in Direct Lake mode, certified datasets with one trusted definition per measure, row-level security, and self-service enablement for business teams.

05

Optimise

Capacity monitoring through the Fabric Capacity Metrics app, cost optimisation, query and refresh tuning, and adoption support, continuing under managed services.

Why Veratas

Why enterprises choose Veratas for Fabric

Fabric is powerful and easy to over-spend on. Veratas brings architecture discipline and real migration experience.

Direct Lake at scale

We build Direct Lake reporting estates that serve thousands of concurrent users, and we know the fallback behaviour and table limits well enough to design around them.

Capacity and cost control

Fabric capacity is metered and shared across workloads. We size the F-SKU against real demand, monitor consumption, and tune workloads so you are not paying for idle compute or hitting throttling.

Governance built in

Microsoft Purview classification, lineage, and access control, plus workspace and domain design, are part of the foundation phase, not retrofitted after the estate is live.

Migration playbooks

Documented playbooks for moving off Azure Synapse and Power BI Premium, including how to handle dedicated SQL pools, pipelines, and existing semantic models.

Proper ALM

Fabric deployment pipelines wired to Git source control, so changes move through development, test, and production in a controlled way rather than being edited live.

FAQ

Frequently asked questions

Quick answers to questions you may have. Can't find what you're looking for? Check out our full documentation.

A foundation plus a first set of workloads typically runs 12 to 20 weeks. Migrations from Azure Synapse or Power BI Premium run 12 to 24 weeks depending on pipeline complexity, the number of semantic models and reports, and how much rework the existing build needs. The assessment phase produces a fixed-price plan, so the range narrows to a committed timeline before delivery starts.
Direct Lake lets a Power BI semantic model read data directly from OneLake Delta tables without importing it and without a live DirectQuery connection to a source database. It gives import-level query speed with no refresh window, because the model sees the Delta tables as they are written. It is the right choice for large, frequently updated datasets, and it is one of the main reasons to put Power BI on Fabric.
For most analytics workloads, yes. Fabric unifies data engineering, warehousing, real-time intelligence, and BI on one platform and one capacity bill, and it removes the operational overhead of sizing and pausing dedicated SQL and Spark pools. That said, Synapse is not end-of-life and some workloads are fine where they are. We run an assessment to confirm the business case, scope, and sequencing before recommending a migration.
Fabric is licensed by capacity. You buy an F-SKU, from F2 upward, measured in capacity units and shared across every workload in the tenant. Content creators and, depending on capacity size, viewers may also need Power BI per-user licences. We size the capacity to your workload, model the licensing, and monitor consumption so the bill matches actual use rather than a guessed starting point.
Both store Delta tables in OneLake and both are queryable in T-SQL. The Lakehouse is built for data engineering: Spark, notebooks, and unstructured or semi-structured data. The Warehouse is a full T-SQL engine with multi-table transactions and a SQL-developer experience. Most estates use both, the Lakehouse for ingestion and transformation, the Warehouse for the curated serving layer. We make that split a deliberate architecture decision.
Yes. The on-premises data gateway connects Fabric pipelines and dataflows to on-premises SQL Server, file shares, and other sources, so a hybrid estate ingests cleanly. For sources that stay in another cloud, OneLake shortcuts can reference Azure Data Lake Storage or Amazon S3 data in place, without physically copying it into Fabric.
A standard Microsoft Fabric foundation (one lakehouse, two pipelines, one report) starts from $10,000. Cost control in Fabric is mostly capacity discipline. We size the F-SKU against modelled workload rather than a round number, monitor consumption through the Fabric Capacity Metrics app, and tune the workloads that drive the most capacity units, typically heavy Spark jobs and inefficient refreshes. Smoothing and bursting absorb short spikes; sustained over-consumption is an engineering signal we act on, not absorb.
At minimum: a workspace and domain topology that maps to your organisation, a clear access model, and Microsoft Purview for data classification, sensitivity labelling, and lineage. Certified semantic models give business users a trusted set of definitions, and deployment pipelines with source control keep changes auditable. We build this in during the foundation phase because retrofitting governance onto a live estate is far harder.
A governed Fabric workspace, one lakehouse with a medallion structure, two data pipelines, a semantic model, one Power BI report, and basic workspace governance. Typically 4 to 6 weeks.
Additional workspaces and source systems, Real-Time Intelligence, a full enterprise warehouse, machine learning, and governance at scale are scoped and quoted separately. Fabric capacity is billed separately by Microsoft.
Get started

Build your data platform on Microsoft Fabric

Whether you are starting fresh on Fabric or migrating off Synapse and Power BI Premium, Veratas brings the architecture and the delivery team. Start with an assessment.

Schedule a discovery call