Capabilities
Six practices on three pillars, delivered by one team that also runs the result
Each practice sits on one of our three pillars. Below is the work we do in each, and the case studies where you can see it.
- Think Data
- Act Analytics
- Perform AI
Data platforms and engineering
Pipelines that pull from many owners' systems every night, reconcile them and keep a history you can query.
What we do
- Ingestion from databases, APIs, files, sensors and the web
- Distributed SQL warehouses and change-only history
- Daily reconciliation between storefront, ERP and partner data
- No-code pipeline builders for analyst teams
Cloud operations and DevOps
We run what we build: containers, delivery pipelines, tested restores and monitoring, in the cloud or in-country.
What we do
- Containerised platforms that move between clouds without a rewrite
- Continuous delivery with reviewed pull requests and secret scanning
- Backups with rehearsed restores, and alerting that reaches a person
- Cost controls such as scheduled shutdown of non-production servers
Geospatial intelligence
Location analysis that turns coordinates into decisions: which road to fix, which courier to use, which branch to price.
What we do
- Linear referencing and hotspot ranking on road networks
- Address-keyword zoning for delivery
- Interactive Arabic maps with right-to-left labels
- Esri geodatabase, GeoJSON and projection handling
Financial and market intelligence
Surveillance and disclosure analytics for supervisors and investors, where every alert shows its evidence.
What we do
- Detection of front-running, pump-and-dump and marking the close
- Market replay to test alert rules before they go live
- Forensic and earnings-quality tests on financial statements
- Checks of press releases against filed statements
Machine learning and entity resolution
Models that match, classify and flag messy real-world records, staged so a person can check every decision.
What we do
- Product and entity matching across Arabic and English sources
- Classification into standard taxonomies such as GS1
- Anomaly and abuse detection with analyst feedback loops
- Statistical models engineers can audit, where the stakes call for it
Generative AI and assistants
Assistants that answer from your own documents in Arabic and English, cite their sources, and report what each answer cost.
What we do
- Retrieval-augmented assistants over large document collections
- Voice agents in Egyptian and Gulf Arabic
- Extraction of tables and figures from PDF filings and reports
- Evaluation sets, latency budgets and cost ledgers per conversation
Sectors we serve
- Government
- National observatories and public consultation analysis, with data kept where the law requires.
- Capital markets
- Surveillance and disclosure analytics that supervisors can explain and defend.
- Retail and e-commerce
- Price intelligence, catalogue matching and the operations data behind margin.
- Healthcare
- Voice AI for clinics, built around patient safety and data-protection approval.
Tell us the decision you need to make.
Describe the decision, the data you hold and your timeline. A senior engineer reads every message.