Data you can trust.Decisions you can stand behind.

I build reliable metrics and data products that give teams a clear basis for decisions.

Permanent role

Lead or Staff team role

Contract work

Focused project or interim lead

Contact me

Worked with

SolarisQontoPentaOptioPayBO1 AGSolarisQontoPentaOptioPayBO1 AGSolarisQontoPentaOptioPayBO1 AGSolarisQontoPentaOptioPayBO1 AGSolarisQontoPentaOptioPayBO1 AGSolarisQontoPentaOptioPayBO1 AG

01Ways to work together

Two ways to work together.

Permanent role

Staff or Lead analytics engineering roles where I can own the foundations and help a team make better decisions.

Discuss a role

Contract engagement

Get experienced help to define, build and validate the metrics your team depends on.

Discuss a project

02Selected work

Built for decisions that matter.

Two projects from one client engagement, followed by public work that shows how I build and think.

Client work · BO1 AG

A North Star framework for clearer product decisions

Connected product priorities to a shared set of growth drivers and more than ten business metrics, giving leadership a clearer view of operational and financial performance.

Designed the KPI framework and the Azure analytics foundations behind it, including ingestion standards, data lake structure and Power BI reporting.

Read the case study

Client work · BO1 AG

Mobile event tracking from definition to deployment

Built an Amplitude analytics framework and mobile event tracking for iOS and Android, covering acquisition, activation, engagement and retention.

Defined the event taxonomy with Engineering, then tested, troubleshot and validated telemetry with QA before deployment.

Read the case study
Bianca Ebert

Client reference

“He was instrumental in taking our data analytics capabilities to the next level. He can not only understand complex data structures, but also prepare them so everyone in the team can work with them.”
Bianca Ebert · CPO, BO1 AG

Public build · Governed Metric Operating Model

Public build · case study

One governed ARR metric

A worked example of defining, testing and publishing one certified metric for consistent use across BI and AI.

Model contracts, independent expected answers, dbt tests and a Snowflake semantic view show how the metric holds together.

Explore the build

Public build · operating model

The operating blueprint for trusted metrics

A broader blueprint for making business metrics owned, validated, published and changeable without losing trust.

Uses the ARR project to distinguish implemented controls from the additional work needed for a production rollout.

Read the blueprint
Explore more writing

03Approach

A clear first result. Then a choice.

We agree what needs to be trusted, make a first result useful, and decide together whether more work is needed.

  1. 01 / Understand

    Agree what needs to be trusted

    We define the decision, the number, its sources and what success would look like.

    A shared definition

  2. 02 / Build

    Build and check a first result

    I build the agreed result and check it against examples your team recognise.

    A result your team can inspect

  3. 03 / Review

    Review it together

    We inspect the result, record what changed and agree whether another step is useful.

    A clear next decision

Timing depends on scope and access. We agree the first milestone before work begins, then choose any further step together.

04Experience

In data since 2016.

10 years across analytics and data engineering, from analyst to team lead, head of insights and staff engineer. M.Sc. Business Intelligence & Process Management, HWR Berlin.

  1. 2024 to now

    Staff Data Engineer

    Consulting

    Governed metrics, data products and AI workflows. Also a primary caregiver.

  2. 2019 to 2023

    Head of Business Insights

    Qonto and Penta

    Built the team and data strategy. Platform testing cut analytics incidents by 30%.

  3. 2016 to 2019

    Analyst to Team Lead

    Solaris and OptioPay

    Built financial data foundations and automated reporting.

Working toolkit

Tools I use

A practical toolkit across analytics engineering, cloud data and applied AI.

Languages

  • Python
  • SQL
  • R

Data engineering

  • dbt
  • Spark / PySpark
  • Pandas
  • Alteryx
  • KNIME

Cloud & warehouse

  • Snowflake
  • AWS
  • Azure
  • Palantir Foundry

Orchestration & CI

  • Airflow
  • Data Factory
  • Jenkins
  • CI/CD

AI & LLM

  • Palantir AIP
  • Pydantic AI
  • LangChain
  • Langfuse
  • Claude Code
  • PyTorch

BI

  • Power BI
  • Tableau
  • Amplitude

05Writing

Engineering notes.

Practical experiments and ideas in applied AI.

All writing

06Contact

Tell me what needs to be trusted.

I am open to permanent roles and contract engagements. Tell me about your team or project, and I will reply within two working days.

Quick brief
What's this about?
Ofer Kulka

© 2026 Ofer Kulka · Data you can trust