I turn raw data into decisions.
6+ years building data functions from the ground up: warehouses, pipelines, governed semantic layers, and production ML & LLM systems for one of the region's largest digital groups.
I build the whole data function: the team, the platform and the roadmap. Then I make sure every number the business sees comes from one governed source.
As Director of Business Intelligence & Data Analytics at Snapp Group, I lead a six-person team covering ETL, warehousing, semantic modeling, BI and production ML/AI. We serve marketing, commercial, operations, finance, production and warehousing.
My background is industrial engineering (M.A., GPA 4.0/4.0), so I think in systems, constraints and throughput. In practice that means I start from the decision someone needs to make and design the data needed for it, rather than the other way round. I'm happiest where data engineering, BI and data science overlap. My current focus is agentic BI: LLM agents that answer business questions against a governed semantic layer.
Four disciplines,
one data function.
Most teams hand work between data engineers, BI developers and data scientists. I've built and led all four disciplines, so the handoffs don't lose anything.
From ten sources
to one source of truth.
The enterprise data platform I designed and run. Hover or tap any component to see its role. Dashboards, ML models and AI agents all read from the same governed semantic layer.
Governed semantic layer
SSAS Tabular holds every KPI definition, business rule and row-level security policy, so a measure means the same thing everywhere it appears.
Projects that moved
real business metrics.
Production systems across marketing, operations, finance and AI. Open any card for the problem, approach, stack and impact.
Don't just read about it.
Play with it.
Small, live versions of patterns I use in production. They run on synthetic data in your browser, so no company data is exposed.
Ask the semantic model
Optimal budget allocation
From analyst
to head of data.
Agency analytics for global FMCG brands, then automotive services, then building and leading the full data function of a large digital group.
Director of BI & Data Analytics
- Lead a six-person team of analysts and engineers across ETL, warehousing, semantic modeling, BI and production ML/AI, including hiring, mentoring and delivery quality.
- Architected the enterprise data warehouse (600 GB+) and ETL platform on Apache Airflow, Python and SSIS, integrating PostgreSQL, MySQL, SQL Server, Google Sheets and web-crawled sources.
- Established organization-wide data governance: data classification, access control, standardized KPI definitions in one semantic layer, data quality standards and RLS.
- Ran the on-premise Power BI Report Server deployment end to end. Tuned indexing and performance to support 5,000+ daily interactions, improving query performance by up to 40%.
- Architected an agentic BI layer in which LLM agents generate DAX against the governed SSAS Tabular model.
- Ran a company-wide AI maturity assessment and co-authored the executive AI roadmap (an internal knowledge agent and a B2B AI product).
- Delivered ML for SKU demand forecasting, marketing optimization, unit economics, throughput and computer vision.
BI Specialist & Data Analyst
- Designed KPIs, BI dashboards and SQL Server reports used in regular strategic reviews with senior management.
- Defined analytical models for product, marketing and operations.
- Automated recurring analytics pipelines to cut manual reporting.
BI Specialist & Data Analyst
- JTI: dashboards and analytics that improved Below-the-Line (BTL) strategy execution.
- Nescafé: market-share analytics for strategic planning.
- Master Foodeh: market analyses and recommendations that improved commercial performance.
Building capability,
not dependency.
Team building & growth
Recruited and developed a six-person team. I run one-to-ones and set individual growth paths, and team members move from junior analysis to owning production systems.
Delivery standards
Introduced structured intake and prioritization, so ad-hoc requests became a managed backlog. Every model or dataset passes peer review and validation before it reaches production.
Executive partnership
Work with C-level and functional heads. I turn ambiguous executive questions into scoped analytical work, and turn the results into decisions leadership can act on.
AI maturity program
Co-own the company's AI maturity program with a product counterpart. It is built on a five-dimension rubric, structured interviews and phased success criteria.
Governance first
Took data governance from nonexistent to an organization-wide standard: one auditable definition per metric, classification policies and role-based access.
Knowledge sharing
Set up a shared channel for AI techniques and prompts. Ideation sessions end as documented decisions, each with a named owner and follow-through.
The stack I
ship with.
Tools I've used in production for years, mostly in high-traffic, on-premise environments where performance matters.
Lessons from
production.
Short write-ups on how I approach data platforms, BI and AI. I share more of these on LinkedIn.
Let's make your
data useful.
I'm open to senior data, BI and analytics leadership roles, including relocation in Europe, and to interesting consulting conversations.