What we do · Eight practices
The team that writes the roadmap builds what it describes.
Eight practices cover the whole job: deciding what is worth building, the data platform under it, the models and the screens people use, the security around them, and the skills that keep it all in use. An engagement draws on whichever the problem needs, and the people who scope it stay to deliver it.
The eight practices
In the order the work runs
Most organizations come in at one stage. Each practice page shows what it delivers, how it starts and the evidence behind it. Where a case study is filed under a practice, its figure is the client’s own.
01
Decide
What is worth building, what it should return, and who answers for it when it is wrong.
Strategy & Governance
Roadmaps, AI governance, NIST RMF
Most AI strategies produce a list. Most governance produces a policy. We turn both into decisions.
Case study15–25% of documents nonstandard — the only ones that now need a reviewerContract compare-and-comply across MSAs, SOWs, POs and invoices
02
Build
The platform underneath, the models on top, and the screens people actually open.
Data Engineering & Platforms
Lakehouse, Snowflake, Databricks
Most data platforms don’t need another rebuild. We keep, fix and build.
Case study30% reporting efficiency gainedA people-data platform over the core people systemAI & Machine Learning
Predictive models, fraud, MLOps
A prediction is not a decision. We build the decision first, then the model.
Case study12% fewer delivery miles per orderAn AI and analytics layer on the existing planning systemGenerative AI & AI Agents
RAG, copilots, agentic automation
Most generative AI pilots never ship. We build the ones that do.
Case study37% fewer Tier 1 live-agent contactsAn omni-channel virtual agent for technical supportBI & Data Visualization
Power BI, Tableau, semantic models
Three reports. Three different numbers. We build the one everyone trusts.
03
Protect
The exposure AI adds, found and closed inside the security program you already run.
AI Security & Cyber Risk
AI attack surface, agent access, cyber risk
AI opened doors your security program was never asked to watch. We find them, rank them and close them.
Case study98% validated first attemptTamper-evident verification of pandemic health events
04
Adopt
The judgment to use what was built, taught role by role and measured by what people do differently.
AI & Data Literacy
Executive training, upskilling
People don’t adopt what they don’t understand. We build the judgment to use AI and data well, role by role.
Case study80% of staff demonstrated improved skillsAn AI and data literacy programme across the workforce
05
Extend your team
For when you know what to build and need senior people in your team to build it.
Staff Augmentation & Expert Services
Senior AI and data specialists who extend your team
Senior data and AI people, embedded in your team. Backed by the people who do the work.
One engagement, two practices
How the practices meet in one piece of work
A global financial information and ratings provider wanted one answer: which vendors were working under current terms, and which were not. Generative AI & AI Agents built the extraction and the comparison against the approved clause library, across MSAs, SOWs, POs and invoices. Strategy & Governance kept attorney review and a full audit trail in the design, so legal could trust the documents it no longer had to read.
15–25% of documents now need a reviewer: the ones that are actually nonstandard. Read the case study
- Contracts, POs, invoices
- Normalize · version
- Extract terms
- Compare to clause library
- Exceptions to counsel
Where to start
Not sure which practice? Start from what you need to decide
Start from a business question
Five questions to answer before the money moves. Most decisions are held up by one of them.GoStart with a fixed-scope engagement
Six bounded first engagements, each with a fixed fee agreed up front and a decision at the end.GoScore where you stand
Four free assessments, a few minutes each. The result is on screen; none of them asks for an email.Go
Every practice works from the same method: Data IQ™ scores where you stand before anything is built.
Tell us what you are trying to decide.
Thirty minutes with a senior practitioner. If none of the eight is the right answer, we will say so.
Not sure which practice? Thirty minutes with a senior practitioner.
