01 / 06
Data Strategy & Architecture
Designing data ecosystems that support analytics, reporting and ML at scale. Clear governance, not just diagrams.
Governance · Lineage · Architecture review
Data & AI consultancy · Abu Dhabi
Strategic data consultancy for complex organisations.
Deep technical expertise, paired with the business judgement to know when and where to apply it: in estates where the data is fragmented, the stakes are regulated, and the answer has to survive an audit.
- Nuclear Energy
- Healthcare
- Government Health
- Insurance & Actuarial
- Technology & FinTech
- 10+
- Years in practice
- 5
- Regulated sectors
- 7
- Client organisations
- GCC
- Primary market
Selected clients
ENECNuclear energy
NMC HealthcareHealthcare
DoH Abu DhabiGovernment health
Dubai Health AuthorityGovernment health
MillimanInsurance & actuarial
Taha ActuariesInsurance & actuarial
Rasan ITTechnology & FinTechServices
What we bring to the table.
Six things an engagement is usually made of. Most start with the first two and end with the last.
02 / 06
ML & AI Advisory
Where to apply ML for actual return, and where not to. Model selection, governance and validation plans.
Model selection · Validation · Governance
03 / 06
Migration & Modernisation
Moving legacy estates onto Snowflake, Databricks or equivalent, without breaking what already works.
Snowflake · Databricks · Cut-over planning
04 / 06
Analytics & Dashboarding
Automated reporting and self-service with Power BI, Tableau, and custom tooling where off-the-shelf fails.
Power BI · Tableau · Custom tooling
05 / 06
Stakeholder Bridging
Translating between executives and engineers. The work that makes the other work possible.
Steering committees · Board packs · Regulators
06 / 06
Team Leadership & Training
Principal-in-residence. Upskilling your teams so the capability outlasts the engagement.
Principal-in-residence · Runbooks · Hiring
Sectors
Industries we’ve served.
Five sectors we’ve delivered work in, past and present, each with a plate drawn from it.
Plate 01Nuclear Energy
ENEC Barakah
Operational analytics and predictive maintenance for the first nuclear energy plant in the Arab world, reported under FANR.
5.6 GW · 4 reactor units · FANR
Plate 02Healthcare
NMC Healthcare
Snowflake migration, clinical standardisation and ML for patient outcomes at the UAE's largest private provider.
10+ hospitals · Snowflake
Plate 03Government Health
DHA · DoH Abu Dhabi
Utilisation analytics, electronic audits and tariff negotiation support for self-insured programmes.
100k+ beneficiaries
Plate 04Insurance & Actuarial
Milliman · Taha Actuaries
GLM reserving, change-point detection and predictive pricing work for GCC insurers and reinsurers.
GLM reserving · IBNR
Plate 05Technology & FinTech
Rasan IT (KSA)
On-premise LLM execution for a FinTech and InsurTech proof of concept, including post-IPO stock analysis.
On-premise LLM · KSA
Methodology
A proven four-phase engagement model.
Each phase closes with something you own: a document, a pipeline, a trained team. Not a status meeting.
Phase 01
Assess
Deep read of the current data estate, governance and business objectives. Quick wins and structural gaps, named separately.
Estate review
Phase 02
Strategise
A tailored roadmap: KPIs, architecture decisions, sequencing and the hiring plan that has to go with them.
Roadmap & KPIs
Phase 03
Implement
Hands-on delivery of pipelines, models and dashboards, with documentation and a handover plan written from day one.
Working systems
Phase 04
Scale
Training, runbooks and hiring, set up so the capability continues without the consultant in the room.
Runbooks & training
- Principal
- Shahzeb Yousuf
- Based
- Abu Dhabi, UAE
- Practice
- Established 2016
- Authorities
- FANR · DoH · DHA
Who you work with
Engagements are led by the person who writes the code.
Every engagement is run by a principal, not scoped by one and then staffed by juniors. The person in the steering meeting is the same person reading your schema, and the recommendation you get has been tested against the data rather than against a template.
The work spans a decade across regulated environments: nuclear operations under FANR, hospital groups migrating live clinical estates, health authorities negotiating tariffs, and insurers whose reserving has to satisfy an auditor. The common thread is that being right is not enough. It also has to be demonstrable.
Data strategySnowflake & DatabricksMachine learningModel governanceRegulated reportingPower BI & TableauTeam leadership
See it in action
The methods, running in your browser.
Interactive demos of the machinery behind the advice. Try them before you talk to us.
Playground demoHow RAG WorksRetrieval-augmented generation over real text: the architecture behind most enterprise LLM systems.Playground demoPredict the Next WordNext-token prediction, visualised: what a language model is actually doing under the hood.Open sourceRUL AnalyzerRemaining-useful-life simulation for a component: the predictive maintenance work behind our energy engagements.Open sourceNRC Technical Letter SearchGrounded question answering over nuclear regulatory reports, with every claim traced to a verified quote.Open sourceFraudLens: Transaction Fraud DetectionReal-time card fraud scoring with SHAP explanations and a full audit trace, built for a bank.All open-source projects →
Contact