Case studies

Engagements that show the range

Advisory and strategy, new platforms, migrations, improving what already exists, agents in production and long-running staffing engagements. From insurance and pensions to manufacturing, energy, software and the public sector.

The range of engagements

Sectors
  • Insurance and pensions
  • Manufacturing
  • Energy and power grids
  • Retail
  • Software and SaaS
  • Public sector
  • Public infrastructure
Technology in the engagements
  • Snowflake
  • Databricks
  • Microsoft Fabric
  • dbt
  • Power BI
  • Azure Synapse
  • Oracle Data Integrator
  • MS SQL Server
Engagement length
  • Four weeks
  • Six months
  • Four years and counting
01 · Advisory and strategy

When the direction has to be set before anything is built

Two pre-studies in the public sector. No platform was built, no code was written. The delivery was decision material.

Municipality · pre-study

From ambition to a decided path

The leadership wanted to become AI-ready and make more decisions on data. What was missing was the path there. The pre-study moved the questions of information ownership and master data to the start of the plan, where they belong.

One quarter, two consultants, eight in-depth interviews. A data vision, a data strategy across six focus areas and a first-year action plan.

Read the case →
Municipal utility · pre-study

From gut feel to data

Investment decisions in the billions were made on the experience of individual employees, statutory reporting was compiled manually, and data sat scattered across dozens of system silos.

Four weeks. A data vision, a data strategy across five focus areas, a three-level KPI architecture and two concrete pilot proposals.

Read the case →
02 · Building and migrating

When something actually has to be built or moved

A new platform from evaluation to production, and a migration where quality had to be proven object by object.

Power grids · new platform

A single view of the grid, in near real time

New grid meters delivered more frequent readings and more types of measurement, but follow-up still happened system by system. Rebtech evaluated Azure, Fabric and Databricks and built the solution on Microsoft Fabric, dbt Cloud and Power BI.

Six months, a data architect and a senior data engineer. Capacity, consumption and production are now followed continuously.

Read the case →
Insurance · migration, under wayWith RAID

From Azure Synapse to Databricks in three months

16 source systems, more than 350 million rows per day and 124 Power BI models, with one hundred per cent report correctness as a contractual requirement. An exhaustive AI-driven assessment took days rather than months.

187 of 400 pipelines were dead and could be cut before the build started. The migration is under way.

Read the case →
03 · Getting more from a living platform

When the solution exists but needs to be faster, steadier or easier to run

One improvement engagement that extended the life of an existing platform, and two agents that have taken over recurring work.

Software · improvement

When the data stopped working overtime

Customers could not begin analysing the month's financial data until late in the afternoon. Rather than a new platform, the heaviest time sinks were prioritised, with a test framework guaranteeing that every improvement produced exactly the same data result.

Load times were cut by up to 55 per cent. At one of the first customers, the time-critical load went from just under twelve hours to a little over five.

Read the case →
Retail · agent in productionWith RAID

The troubleshooting is done before anyone logs in

The nightly load fails at quarter past two. At six, an agent reads the logs and determines where the fault belongs. Six outcomes are defined, and an ordinary script decides which one it was.

The diagnosis is done before the working day begins.

Read the case →
Retail · agent in productionWith RAID

Documentation became a review of about ten minutes a week

Around 950 dbt models and a dozen developers in the same repo. Every Sunday, an agent compares the documentation against the code and leaves a finished change proposal.

If it cannot evidence what holds today, the document is marked uncertain.

Read the case →
04 · Staff augmentation

When you own the engagement and we staff it

The longest collaboration we have, and the one that best shows what the nearshore network can do.

Manufacturing · four years and counting

From staff augmentation to strategic partner

The client could find neither the right expertise nor the right price in Sweden. Rebtech activated its European network and staffed the engagement with an ODI specialist and a Power BI specialist. When the client later migrated to Snowflake, the consultants became key people in the migration. When the product owner moved on to a new internal role, he explicitly asked to take Rebtech's consultants with him.

Since autumn 2022. The engagement has grown from two to four specialists, with delivery from Poland and Romania.

Read the case →
Next step

Do any of them sound like you?

Book a meeting →Book a meeting