Senior Data Scientist / Data Analyst Team Lead

3/2022 – 10/2022

Responsibilities:
- Owned and managed the data lake in BigQuery, housing customers' carbon footprint data.
- Designed the data structure in the data lake for different types of data entities.
- Implemented effective data management rules and processes.
- Enhanced data sourcing and ETL processes for semi-structured data ingestion.
- Defined the dashboarding process and developed interactive dashboards.

Major accomplishments:
- Collaborated in the design and implementation of ETL processes and data infrastructure for an early-stage start-up.
- Consolidated diverse and semi-structured data inputs into a unified dataset and created standardized, user-friendly dashboards for customers.

Tech stack: Python, BigQuery, MySQL, Google Data Studio, Kubernetes

SQL, Python, ETL

9/2019 – 2/2022

Responsibilities:
- Developed and implemented new fraud detection rules for mobile attribution.
- Reviewed and optimized the performance of existing rules, ensuring their effectiveness.
- Led cross-functional teams and reported to senior stakeholders for the largest projects.
- Conducted extensive data mining, analyzing billions of daily data points across diverse datasets (ad views, clicks, installs, in-app behavior) to identify fraudulent patterns.
- Utilized statistical properties, distribution analysis, time-series modeling, and white-box AI algorithms for accurate fraud classification.
- Acted as a key point of contact for customers and partners, collaborating on fraud cases.
- Collaborated with product managers, front-end and back-end developers to implement updates to the fraud detection system.
- Created and maintained dashboards for internal and external stakeholders.

Major accomplishments:
- Successfully deployed a critical fraud detection algorithm (Bayesian network) resulting in a 30% increase in fraud identification with high accuracy. Oversaw the end-to-end implementation of this substantial project involving multiple teams.
- Revamped data aggregation approaches and rewrote numerous rules, leading to a 10% increase in fraudulent activity detection.
- Defined a blueprint for an enhanced and flexible fraud detection architecture, enabling advanced rule creation through expanded data aggregation capabilities.

Tech stack: Bigquery, Python, Looker, Tableau, Airflow, Scala on Spark

Data Science, Data Mining, SQL, Python

3/2017 – 7/2019

Responsibilities:
- Designed and developed a fully functional prototype for a product that converts images of charts and graphs into CSV files.
- Programmed the generation of hundreds of thousands of images and their metadata to train AI algorithms effectively.
- Developed and trained multiple neural network-based AI models for precise object detection and image classification.
- Constructed an efficient architecture comprising parallel running multi-tasking processes synchronized through queues and a backend database.

Major Accomplishments:
- Single-handedly conceptualized and built a fully operational prototype from the ground up.
- Applied an iterative approach to prototype development, scaling the product incrementally for optimized efficiency.
- Overcame steep learning curves in Python programming, AI techniques, and cloud architecture to master complex technologies effectively.

Tech stack: Python, Keras, Neural Networks, scikit-learn, YOLOv2, AWS, MySQL, VBA, Excel

SQL, Faltendes Neuronales Netzwerk (CNN), Keras, Python