Data platform assessment and roadmap
Find out whether your data is ready for the reports and the AI you want, and leave with an architecture, a budget and a roadmap you can take to your board.
Know what your data can support before you buy anything
The most expensive data platform mistakes happen before anything is installed: a warehouse bought for reports nobody reconciles, an AI pilot built on documents nobody is allowed to share, a vector database sized for a demo, or a vendor platform chosen because its brochure matched the tender.
Our assessment answers the questions first, with evidence. We look at the data you hold, the pipelines that move it, the people who own it and the rules you work under, then recommend an architecture, a route, a budget and a roadmap that fit them, ordered so the first use case pays for the next.
When it fits
01
An AI programme is starting
The board has approved AI, and nobody can yet say which data it will use or whether that data is fit for it.
02
Reports take longer every year
Regulatory returns and management reports are assembled by hand from extracts, and the numbers disagree.
03
A renewal or refresh is due
A warehouse, database or storage contract is coming up, and the platform choice changes the budget.
04
Vendors are pitching platforms
Dell, HPE, Oracle and the cloud providers all have a data platform to sell, and you want an independent comparison.
What we do
Data readiness review
Score your data on six measures: quality, freshness, lineage, access control, skills and governance, with the evidence behind each score.
Source and pipeline inventory
Map the systems that hold data, the pipelines and extracts that move it, who owns each one, and where the same data is copied twice.
Use-case selection
Choose the first reporting and AI use cases on value, data readiness and risk, with a measurable result for each.
Target architecture and route
Recommend the platform route, open source on OpenShift, a vendor data platform, Oracle-centred or hybrid, and draw the target architecture.
Sizing and cost
Size storage, compute, GPUs and licences from your actual data volumes, and model each route's cost over three to five years.
Roadmap and business case
A phased plan from the first pipeline to the full platform, with milestones, risks and the benefits that justify each step.
Data readiness review
Six measures, scored today and set as targets
We score each measure from interviews, configuration reviews and samples of the data itself. The gap between today and the target on each row becomes the roadmap, ordered to remove the biggest risks to your first use cases first.
- Today
- Where each measure stands now, from evidence rather than opinion.
- Target
- Where it needs to be for the use cases you have chosen, agreed with you.
- Six measures
- Quality, freshness, lineage, access control, skills and governance.
How an assessment runs
01
Discover
Workshops with business, data, IT, security and compliance teams, and a tour of the systems and reports you run today.
02
Sample and score
Profile real data for duplicates, gaps and freshness, trace a few key numbers to their source, and score the six measures.
03
Design
Draw the target architecture, compare the routes on cost, risk and fit, and size the platform from your volumes.
04
Plan
Agree the first use cases, a phased roadmap, a budget and a business case, ready for approval and procurement.
What you get
- A data readiness score across six measures, with the evidence behind it
- An inventory of sources, pipelines, extracts and owners
- The first reporting and AI use cases, each with a measurable result
- A platform recommendation, with the routes compared
- A target architecture for ingestion, storage, lakehouse, AI data and DC/DR
- Sizing for storage, compute, GPUs and licences, with cost models over three to five years
- A phased roadmap and business case
What we compare
Each route is weighed on your data, skills, rules and budget.
Routes
- Open source on Red Hat OpenShift
- Dell AI Data Platform
- HPE unified data layer
- NetApp AFX and AI Data Engine
- Oracle Database and Exadata
- Hybrid with a public cloud
What we look at
- Core banking and ERP extracts
- Oracle and PostgreSQL databases
- Document archives and scans
- Logs and sensor data
- Existing warehouses and BI
- Regulatory and audit rules
Questions about assessments
How long does an assessment take?
Weeks, not months. The length depends on how many systems, reports and teams are in scope, and we agree it with you at the start.
Do you need access to our data?
We need samples, under your rules. Profiling real records is what turns opinions about data quality into evidence. We work on your premises, on anonymised extracts where required, and never take data out of the building.
Will you recommend the vendor you partner with?
We partner with Red Hat, Dell, HPE, Oracle and NVIDIA, and we also build on open source. The recommendation follows the evidence on your data, rules, skills and budget, and we show you the comparison so you can check it.
What happens after the assessment?
You own the architecture and the roadmap, and can take them to any supplier. Many clients ask us to build the platform, which we run as a separate project.
Architecture and designAssess your data with evidence
Tell us what you want the data to do and the rules around it. We'll propose the scope of an assessment and what you will have at the end of it.