Led a team of 4 people towards a rebuild of a serverless OCR application in a high visibility environment with intense stakeholder attention. Handed over the project to an outsource team after delivering the MVP. Technologies used: Claude 3, Python, Textract, Lambda, S3, SQS/SNS, Terraform and GitLab.

For more info see: https://www.perceptivebits.com/

Built a new highly modular Everything-as-Code Azure data platform (called SIDAP) with an excellent team. Completely automated CI/CD in Azure Devops. The platform has different subscriptions per DTAP environment and every end-user group also having its DTAP environment subscriptions. Separate Terraform statefiles for modular (and faster) deployment. Processing is done via PySpark in Databricks workspaces, source meta data is configured in YAML contracts. My activities encompassed:-Built and mentored a young talented team and leaving them with a developer centric culture. - Was involved in assessing and hiring most of the platform and data engineers.- End-responsible for the solution architecture of the platform. Designed the solution together with the team including infrastructure, data flow and ETL algorithms (e.g. deduplication and Change Data Capture in PySpark).- Designed and built a data quality engine called Estella based on the Great Expectations framework- Guided the software architecture and developed features for the Azure Data Factory pipeline generation, PySpark based ETL framework, the Terraform modules and the data quality engine.- Introduced and pursued a Domain Driven Design way-of-working

As a hands-on interim CTO I was part of the MT and helped SphereMall improve on different levels.On a strategic level I helped steer the company by planning features for all products, resources and customer requests, while being the linking pin between Business and Tech. On a human resource level I developed the team towards the right skill set, coached the team to become more value oriented, recruit the appropriate next roles, and push the culture towards knowledge sharing and taking ownership of problems.On an architectural level I removed complexity both in process, cloud components and code over all AWS projects, while also restructuring the applications to interact to a central data foundation. Way of working was improved by introduction of separate AWS environments for testing, switching to central user management and Single Sign-On, and measuring code quality through SonarCloud, code reviews and pull-requests. CI/CD was made to follow an appropriate adaptation of GitFlow.On a design level I closed the gap between Business and Tech by introducing Domain Driven Design, thereby facilitating a common language for the whole company. Personally I designed (functional and technical) several new features, including the Customer Journey Tool, the Nucleus Data Foundation, Global SphereMall Entity Identifier and the URL Mapping Tool.Finally, on a code level, I was directly involved with building and deploying new components and features for their product suite, including a Snowplow implementation on multiple channels, a Spark based Redshift to Neptune ETL, several Python and TypeScript Serverless services and the implementation of the above mentioned designs.I would like to thank SphereMall for an awesome project, being a great client to work for and just overall being fun, professional and success oriented.

Lead the organization to become a recognized thought leader on data, software and AI. Greatly increased innovation by promoting knowledge sharing on all levels.Responsible for all technical expertise at Anchormen, headed the Data Science and Engineering departments (~ 55 people). Incubated both the MRO and Data Hub solutions, which are now successfully implemented at numerous clients.Built Anchormen’s national image as a high caliber consultancy through public speaking, organizing meetups and hosting the AI Journey podcast.Improved quality of work by fostering a culture of sound software development.Made Anchormen an attractive company to work at, and hired ~50 new engineers and scientists.

Made our clients more data driven and cost-effective. Projects included the design and implementation of data solutions e.g. recommendation engines, predictive maintenance, sentiment analysis, precision marketing.

A startup that accelerated and improved business strategy development. A platform which detects accelerating innovation signals on the web and provides tools for qualitative analysis of those signals. The platform consisted of a Python micro-service architecture, with a NumPy based topic-model (nHDP) and a Titan-Hbase graph backend.

Allowed a big retailer to revolutionize their E-Commerce platform. Team lead for a service development team which implemented several services in a SOA environment. Implemented a high performance pricing engine with hot-swappable calculation schemas from SAP.

Became Java Lead of Seafight in April 2012. Greatly extended the lifetime of a 6 year old project by reducing technical debt and bringing new features. Team lead of 6 people. Introduced Guice, EasyMock unit testing, code reviews, continuous integration, GIT and an effective branching and release strategy. Implemented DB caching, a new AI, flexible and easily extendable XML based loot distribution strategies and more in a highly concurrent environment.

Became Dev Lead of Kultan in December 2012Main achievement: Implemented a distributed persistent guild system using Eclipse Link and ActiveMQ and created a 'winter setting' in the client using shaders.

Developed a website for a crossword puzzle manufacturer.

Developed location based services using Google Maps.

Team size of approximately 50 telemarketing agents