Skip to content
BI / Data / Analytics Lead

Data and reports
that leadership relies on.

I design data platforms (Snowflake, data mesh on Azure) and Power BI models. I merge fragmented reports into one model with agreed metric definitions, which finance and operations teams use on their own.

scroll
source consolidation · sample data
ExcelCSVSQLAPIERPe-mail → one model
24 scattered sources → 1 source of truth
01Flow

From sector to capabilities

The chart links sectors to projects and projects to capability groups. One unit of ribbon width is one stack item from a project description.

Point to a node or ribbon to trace its path. Click a project to go to its description.
FMCG / E-COMMERCE2 projects FMCG / FINANCE1 project GENAI / ENABLEMENT1 project BANKING1 project FINANCIAL CRIME / AML1 project 01↓Consolidating fragmented reporting inSnowflake 02↓Data mesh and domain self-serve 03↓Power BI for hundreds of recipientsand accurate variable pay 04↓Piloting a knowledge agent foranalysts 05↓Transaction-code integrity controls 06↓Fewer false positives in transactionmonitoring Data platform4 links Semantic modeling1 link Governance & quality4 links Reporting & UX1 link AI & automation4 links Tools1 link Banking & financial crime3 links FMCG / E-COMMERCE FMCG / FINANCE GENAI / ENABLEMENT BANKING FINANCIAL CRIME / AML 01↓Consolidating fragmented reportingin Snowflake 02↓Data mesh and domain self-serve 03↓Power BI for hundreds ofrecipients and accurate variable… 04↓Piloting a knowledge agent foranalysts 05↓Transaction-code integritycontrols 06↓Fewer false positives intransaction monitoring Data platform Semantic modeling Governance & quality Reporting & UX AI & automation Tools Banking & financial crime FMCG / E-COMMERCE FMCG / FINANCE GENAI / ENABLEMENT BANKING FINANCIAL CRIME / AML 01 02 03 04 05 06 Data platform Semantic modeling Governance & quality Reporting & UX AI & automation Tools Banking & financial crime
  1. 01Consolidating fragmented reporting in Snowflake
  2. 02Data mesh and domain self-serve
  3. 03Power BI for hundreds of recipients and accurate variable pay
  4. 04Piloting a knowledge agent for analysts
  5. 05Transaction-code integrity controls
  6. 06Fewer false positives in transaction monitoring
The data comes from the project descriptions and the capability list on this page.
02Work

Selected projects: the problem, what I did, and the result.

FMCG / E-commerce

Consolidating fragmented reporting in Snowflake

P
Reporting was spread across many inconsistent sources, with no shared metric definitions.
A
I led the migration to Snowflake from start to finish and moved reporting onto one governed layer.
R
Every team calculates KPIs from the same definitions.
SnowflakeELTKPI governance
FMCG / E-commerce

Data mesh and domain self-serve

P
The central data team was a bottleneck for a growing number of business requests.
A
I designed a data mesh architecture on Azure in which business domains get data as a product.
R
Domains work on their own, and governance rules and quality checks stayed in place.
AzureData meshData quality
FMCG / Finance

Power BI for hundreds of recipients and accurate variable pay

P
Hundreds of recipients and executives needed reliable dashboards, and variable-pay settlement was error-prone.
A
I built automated financial and sales dashboards with per-region RLS, plus a variable-pay reconciliation process.
R
~600 recipients and ~50 HQ reports on one standard. A reconciliation step keeps variable-pay payouts accurate.
Power BIDAX / RLSReconciliation
GenAI / Enablement

Piloting a knowledge agent for analysts

P
Report logic and business rules lived in people’s heads, so onboarding took longer.
A
I work with GenAI daily (Power BI Copilot) and am piloting a Copilot 365 knowledge agent that captures report logic.
R
Pilot in progress. Goal: report knowledge on demand and faster onboarding.
Power BI CopilotCopilot 365MCP
Banking

Transaction-code integrity controls

P
Transaction codes had to stay consistent across global systems.
A
I built controls and tooling in MS Access and VBA that check transaction-code integrity automatically.
R
Code integrity is checked automatically and continuously.
MS AccessVBAIntegrity controls
Financial Crime / AML

Fewer false positives in transaction monitoring

P
Monitoring scenarios generated too many false positives, which overloaded the team.
A
I tuned AML scenarios (detection indicators and parameters) for precision.
R
Fewer false positives, so the team spends its time on cases that need investigation.
AMLTransaction monitoringScenario tuning
03Capabilities

Areas and tools I work with.

model
DIM_Category CategoryKey (PK)Name · Group DIM_Segment SegmentKey (PK)Tier FACT_metrics ∗ CustomerKey (FK) ∗ ProductKey (FK) ∗ DateKey (FK) Amount · Quantity DIM_Date DateKey (PK)Year · Month DIM_Customer CustomerKey (PK)Segment DIM_Product ProductKey (PK)Category DIM_Channel ChannelKey (PK)Type DIM_Region RegionKey (PK)Country
A star schema: one fact table and shared dimensions. It keeps a semantic model fast and easy to maintain.
Revenue YoY %.dax
Revenue YoY % =
VAR _cur = [Total Revenue]
VAR _prior =
    CALCULATE (
        [Total Revenue],
        SAMEPERIODLASTYEAR ( 'Date'[Date] )
    )
RETURN
    DIVIDE ( _cur - _prior, _prior )
DAX in practice: measures built with VAR/RETURN evaluate each expression once, so they run faster and read better.
Data platform
SnowflakeAzure (data mesh)ETL / ELTSQLLayered architecture
Semantic modeling
Dimensional modelingDAXPower Query (M)Tabular EditorRLSPerformance tuning
Governance & quality
KPI governanceData qualityReconciliationDocumentation
Reporting & UX
Power BIFinance/sales dashboardsExecutive reportingFigma (UX)
AI & automation
GenAI / LLMPower BI CopilotCopilot 365MCPPower AutomateAlteryxVBA
Tools
Advanced ExcelMS AccessPythonJiraAgile
Banking & financial crime
AMLTransaction monitoringScenario tuning
04By the numbers

The scale I work at

Real figures, rounded.

EXP
~10 yrs
in BI and data analytics
TEAM
10 people
in the analytics team I led
REACH
~600
dashboard recipients in one organization
EXEC
~50
reports for executives and HQ
05Process

How I work, from the first business question to a finished report.

01

Decision first, data second

I start with the question the report has to answer, and only then look at what data exists.

02

A shared data layer

I bring data from many sources into one model. Each metric has one definition, and the access and quality rules are clear.

03

Performance and correctness

I tune models so reports are fast and the numbers reconcile to the cent.

04

Self-serve and documentation

Teams get tools and documentation, so simple questions don’t wait in the data team’s queue.

06About

I’ve spent about ten years in BI and data analytics, in banking, fintech, FMCG and e-commerce. I design data architecture and I’m responsible for the numbers in reports being right and the reports being used.

I led a ten-person analytics team working in Agile. We worked with Finance, Sales and the executive team. I owned a migration to Snowflake, a data mesh design on Azure, and Power BI rollouts for hundreds of recipients.

I use GenAI daily in analytical work. I run Power BI training (previously Excel) and explain complex data topics to the people who make decisions based on them.

Contact

Let’s talk data.

If you want to sort out your analytics or need a BI lead for your team, send an email or a LinkedIn message.

Available for senior projects and roles