About
Carsten Böhm
Data engineer and data lead based in Dubai. Currently running data and analytics at a UAE payments fintech, with eight years of hands-on experience across IoT, streaming, and analytical platforms.
Background
The story
I spent five and a half years at Alexander Thamm GmbH, Germany's biggest data consultancy, ending as Principal Data Engineer. I led IoT pipelines processing logs from 4 million cooking devices, built a recipe recommendation system serving 30 million requests a day for Vorwerk's Thermomix, and shipped streaming pipelines for food and beverage processing.
In late 2023 I moved to Dubai and joined Pyypl, a UAE payments fintech, as Head of Data & Analytics. I spearheaded a Databricks migration that cut ETL costs by half, re-engineered pipelines for real-time transaction processing, and built out the reconciliation data models the finance team now runs their day on. That engagement continues now as a contractor.
Along the way I watched a lot of teams get sold tools they didn't need. I've seen companies spend $50k a year on Snowflake when PostgreSQL would have been fine. I've seen modern data stacks with 12 tools where 4 would have done. What I try to do now is the opposite: understand what you actually need and help you ship it without enterprise overhead you can't afford yet.
Philosophy
What I believe
Start simple, scale when needed
The best architecture is the simplest one that solves your current problem. Complexity should be earned through actual scale requirements, not anticipated based on ambition.
Vendor-neutral recommendations
I don't have partnerships or referral agreements with vendors. My recommendations are based on what works for your situation, not what pays me the most.
Implementation matters more than slides
I'm a practitioner, not a presenter. I've built the systems I recommend. Strategy is valuable, but only if it leads to working infrastructure.
Your team should own it
I build systems your team can maintain. Full documentation, knowledge transfer, and patterns they can extend. Not black boxes that require me to keep running.
Experience
Selected work
Current engagement
Data lead at a UAE payments fintech
Running the data and analytics function at Pyypl since late 2023, originally as Head of Data & Analytics and currently as a contractor. The mandate is simple: turn transaction data into decisions the business can act on, keep the platform reliable under real-time load, and make the finance team's work easier every quarter.
BI & decisions
Dashboards for finance, ops, product, and leadership, built around the questions each team actually asks. KPIs they trust because the underlying models are documented and owned.
Finance automation
Reconciliation data models that replaced spreadsheet work with pipelines finance can run and triage themselves. Exceptions surface automatically instead of being hunted down at month-end.
Platform economics
Migrated the data platform onto Databricks and cut ETL processing cost by 50% while moving to real-time transaction ingestion. Lower bill, faster insights, room to scale.
Before Pyypl, five years at Alexander Thamm GmbH in Munich, Germany's biggest data consultancy, ending as Principal Data Engineer.
4M+
devices
IoT pipelines for connected cooking devices
ETL on semi-structured logs into analytics-ready tables, powering personalised recommendations, error analysis, and management dashboards.
30M/day
requests
Recipe recommender for Vorwerk's Thermomix
End-to-end recommendation pipeline and API serving the Cookidoo platform. Led a team of three engineers.
Real-time
streaming
Streaming IoT for food and beverage processing
Real-time pipelines turning machine messages into operational KPIs for a leading equipment supplier's customers.
Stack
Technologies I work with
Location
Based in Dubai
Operating from Dubai, fully remote. I work with clients across multiple time zones.
Primary regions: Middle East, DACH (Germany, Austria, Switzerland), UK, and broader EU.
Working hours coverage
- Middle East: Full overlap
- Europe: Strong overlap (morning + afternoon)
- UK: Good overlap (afternoon)
- US East: Limited overlap (early morning)
Let's talk about your data challenges
No sales pitch. Just an honest conversation about what you need.