Beavertown Brewery
Wrapping around a nascent data team as it built its cloud data platform.
Beavertown Brewery is one of the UK's most distinctive craft breweries. As the business scaled, it needed a modern data capability: a small hub-and-spoke data team, supported by a cloud data platform built on Azure Databricks and Power BI.
The team was new and growing, training up on cloud data engineering while migrating off a legacy reporting tool. At the same time they were standing up a Microsoft-based platform inside a parent group's tenant and delivering value to colleagues in sales, finance, operations, purchasing and marketing.
We looked at a few consultancies. We didn't want someone to come in, deliver a product and walk away. The Brain Trust engagement is the opposite of that. Like a wrap-around or bolt-on to an existing data team to broaden its horizons, deepen its experience and raise the bar of its ambitions.
Beavertown turned to endjin's Brain Trust engagement to support the internal data team during this formative period. The Brain Trust ("Brain Power, by the Hour") gives on-demand access to endjin's experts across data engineering, data architecture, Power BI and data strategy. Sessions ran on a fortnightly cadence and flexed between high-level strategy and hands-on technical deep dives, with the agenda shaped around whatever the team needed to unblock next.
Solid Foundations
Foundations came first. Endjin helped the team establish the engineering practices that let a hub-and-spoke data function move confidently from day one: Azure environment topology across Dev, Test, Acceptance and Production, infrastructure-as-code, an Azure DevOps branch strategy, gated deployments, observability, and a shared definition of done. Alongside this, sessions shaped a Power BI architecture aligned to audiences and the decisions they needed to make.
That meant a domain-based workspace topology for sales, finance, operations, purchasing, inventory and production, clear separation between modellers and reporters, and deployment pipelines promoting artefacts through each environment. The legacy reporting tool's single monolithic model gave way to a set of focused semantic models built on a medallion architecture and classic star-schema principles.
Bolder Ambitions
With foundations in place, attention turned to hardening the platform for a business carrying sensitive commercial and finance data inside a larger parent group. Endjin worked with the team on a layered security model: role-based access via Entra ID groups, row-level and object-level security in Power BI, dynamic row-level security driven by the logged-in user, and a YAML-driven approach to managing Databricks SQL warehouse access as code.
On governance, the Brain Trust helped the team work through a busy field of options (SharePoint lists, Microsoft Purview, Databricks Unity Catalog, third-party catalogues) and choose the right level of investment for its stage, supported by an Architecture Decision Record framework that captured objectives, constraints and trade-offs in a consistent format.
The Brain Trust also stretched beyond Databricks and Power BI where the business needed it to. Fragile, SAP-connected Excel workflows were replaced with Dataverse-backed Power Apps, applying the same environment and deployment muscles in a different part of the Microsoft stack. The harder single-customer-view problem was approached as a data-science experiment: data contracts and cleaning first, then deterministic matching on strong identifiers, before layering in probabilistic matching with confidence scoring and a human in the loop.
Throughout, endjin helped the team derive its data and analytics strategy from the wider company strategy, so every initiative could be justified by the audiences and decisions it served.
The Brain Trust engagement has been incredibly valuable for the Beavertown data team: endjin supported us in setting up a nascent hub and spoke data team at Beavertown and migrating to Azure Databricks and Power BI. Impactful sessions with the team covered everything from DataOps best practises and ways of working, to designing our Power BI architecture, to the more holistic data & analytics strategy and laddering that down from company strategy.
Alongside the technical work, the Brain Trust has acted as an independent sounding board for the team, benchmarking their approach against the wider industry and challenging internal assumptions where it helped. It has also helped them work through the practicalities of operating inside a parent group's shared infrastructure.
Beavertown has already gone live with four core Power BI apps covering sales, finance, operations and purchasing. They sit on a disciplined deployment pipeline and a clear security model, run by a small but capable team that has grown into the new platform.
The Brain Trust engagement continues, with endjin mentoring the team as the platform evolves and the company's data ambitions grow.
