Senior Data Scientist · AI & Data Engineer

Lead data scientist on the fraud-detection platform, owning ML systems end-to-end — from cloud data pipelines to Agentic AI investigation tools running in production.

• Architect and operate real-time fraud-detection systems that score large volumes of retail transactions, combining gradient-boosted models, anomaly detection, and graph analytics to stay ahead of evolving fraud patterns.
• Build Agentic AI and LLM-powered investigation copilots with RAG and GraphRAG over a Neo4j knowledge graph — automating alert triage, compressing investigation time, and exposing organized fraud rings.
• Apply Causal Inference to separate genuine fraud drivers from spurious correlations, sharpening intervention targeting and reducing false positives.
• Engineer end-to-end big-data and ML pipelines on the cloud with Databricks & Apache Spark (Azure Data Factory, Synapse, Data Lake), backed by production MLOps — CI/CD, feature stores, model monitoring, and drift detection.
• Partner with fraud analysts, engineering, and compliance to turn domain expertise into features and feedback loops that continuously strengthen detection coverage and trust.

Senior Data Scientist & Computer Vision Expert

Project: Cidaas-ID-Validator

2018 - 2023

• Led the design and deployment of scalable microservices that productionized ML models for identity verification, improving throughput and responsiveness.
• Developed a high-precision OCR model for extracting structured data from diverse identity documents, raising data-capture accuracy.
• Built a security-check framework combining Computer Vision and Probabilistic Graphical Models to reliably detect forged and manipulated documents.
• Created an eye-tracking-based Human-Computer Interaction module (ninoxipy) for real-time liveness detection, strengthening anti-spoofing defenses.
• Established the team's MLOps framework for model integration, monitoring, and lifecycle management, enabling continuous and reliable delivery.

Expert Data Scientist

Project: Cidaas-FDS (Fraud Detection System)

2020 - 2023

• Architected an ML-driven cybersecurity solution that embedded fraud detection into the platform core.
• Built RESTful APIs serving fraud models for real-time scoring and immediate response to security threats.
• Designed and maintained the CI/CD pipeline for automated testing and agile model deployment.
• Developed a Smart Multi-Factor Authentication (MFA) system using a Fuzzy Inference System that adapts verification to real-time risk levels.
• Ran Causal Inference analyses to pinpoint the root causes of fraud and drive targeted prevention.
• Reached up to 97% fraud-detection accuracy, materially reducing fraud risk and its impact on the platform.

Senior Data Scientist

Project: Bosch-EBR (Experience-based Repair)

2020 - 2023

• Built an NLP-based vehicle-identification system using Named Entity Recognition (NER) and graph theory to accurately classify vehicle components.
• Delivered the model through a RESTful API for seamless stakeholder integration and accessibility.
• Engineered a resilient stream-processing framework with Faust for EBR microservices, improving scalability and fault tolerance.
• Created an LLM-powered Concept Identification Service that mines automotive forums for failure discussions and links reported problems to validated repairs — accelerating diagnosis and resolution.

Lead Data Scientist

Project: Porsche-CCD (Corner Case Detection)

2018 - 2019

• Engineered a spatio-temporal model fusing onboard sensor data and vehicle geolocation to detect driving corner cases, strengthening predictive safety.
• Designed an interactive React monitoring UI for real-time fleet visualization and corner-case identification.
• Built a high-throughput streaming system with MQTT and Tornado for reliable real-time data ingestion and processing.
• Delivered rich interactive analytics with Plotly and Bokeh to support faster, data-driven decisions.

Computer Vision Expert & Data Scientist

Project: Chinese Painting Seal Assessment

2018 - 2019

• Developed a CNN model in TensorFlow to identify Chinese artists' seals, enabling high-precision artwork authentication.
• Built a recurrent neural network to assess seal originality and distinguish genuine works from forgeries, supporting cultural-heritage preservation and art-market integrity.

AI Researcher & Developer

• Modeled human gaze behavior in the Immersive Quantified Learning lab (iQL), advancing research in Human-Document Interaction.
• Designed an eye-tracking-based HCI system for the AICASys project, improving digital accessibility for users with disabilities.
• Developed a generative probabilistic model for synthesizing eye-movement patterns to better understand and predict user interaction.
• Served as Teaching Assistant for the Data Mining course at TU Kaiserslautern.
• Supervised Master's and Bachelor's theses in AI and HCI, mentoring students from concept to defense.

Full-Stack Developer

• Delivered an organizational personality-assessment platform for Mapna Group, deployed to 20,000+ employees to support workforce management and development.
• Built the full stack end-to-end — PHP back-end logic with Python-based data analysis and processing — taking the assessment portal from concept to production.

Computer Science Lecturer

• Taught Machine Learning, Data Engineering, and Computer Vision to undergraduate Computer Science students, pairing strong theoretical foundations with hands-on, real-world application.
• Taught Data Analysis and Advanced Statistics to postgraduate Clinical Psychology students, equipping them with rigorous quantitative tools for research and analysis.