Senior Software Engineer at Leobit

Built AI-powered APIs and SDKs exposing trained models as developer-friendly services, enabling seamless integration of intelligent features into customer applications. Developed real-time inference services for LLMs and Transformer models, optimizing request handling, caching, and batching for high-throughput production workloads. Integrated vector search engines such as Weaviate and pgvector for semantic retrieval and contextual chat. Implemented agentic workflows using LangChain for orchestrating multi-step reasoning tasks, powering advanced automation. Extended applications with AI features such as summarization, document classification, and conversational assistants. Designed modular components for authentication, monitoring, and error recovery to enable enterprise-scale deployment. Optimized model inference with quantization, pruning, and GPU acceleration for latency-sensitive environments like chat and recommendations. Collaborated with product teams to translate user-facing req

Machine Learning Engineer at Zalando

Developed and deployed computer vision services classifying millions of product images for production-ready REST/gRPC APIs powering catalog enrichment and search improvements. Engineered real-time recommendation APIs processing live interaction streams for personalized suggestions across mobile and web with low latency. Integrated graph-based search and anomaly detection into fraud-prevention systems, packaged as reusable components. Built feature-flagged A/B testing infrastructure for AI-powered personalization, enabling controlled experimentation and safe rollout. Designed and deployed training-to-inference workflows on AWS SageMaker, exposing models as microservices and integrating into user shopping experiences. Implemented demand forecasting APIs using time-series models to support supply chain and inventory optimization. Collaborated with frontend/backend teams to embed ML APIs into web and mobile flows for seamless UX. Delivered tooling for dataset versioning, feature encoding,

AI Specialist at Capgemini

Delivered AI-enabled applications for financial, healthcare, and automotive clients by packaging ML algorithms into production-ready microservices that could be embedded into enterprise workflows. Built multilingual NLP services for customer service automation, exposing chatbot APIs integrated with CRMs, reducing manual case handling while preserving brand consistency. Developed predictive maintenance APIs processing sensor and IoT data in real time, enabling proactive equipment monitoring and direct integration into industrial dashboards. Migrated on-prem ML workloads to Azure ML and Databricks, transforming ad-hoc scripts into standardized, reusable pipelines. Implemented ensemble-based forecasting services deployed as backend components providing stable demand predictions for client planning tools. Built dashboards and APIs translating AI outputs into actionable metrics for executives. Integrated deep learning architectures like CNNs/RNNs for real-time voice, image, and document str

Data Scientist at Nortal

Developed early neural network implementations for industrial automation, optimizing sensor-driven data streams for efficiency. Led feature engineering, clustering, ETL, data warehousing, validation studies, and stress testing to assess model robustness in high-risk environments. Authored internal best-practice documentation influencing company-wide adoption of data-driven methods in industrial innovation programs. Implemented scalable ETL processes integrated with enterprise data warehouses; designed dashboards translating AI outputs into simple, actionable metrics for executives.

Senior Software Engineer at Leobit

Built AI-powered APIs and SDKs exposing trained models as developer-friendly services, enabling seamless integration of intelligent features into customer applications. Implemented real-time inference services for LLMs and Transformer models, optimizing request handling, caching, and batching for high-throughput production workloads. Integrated vector search engines (Weaviate, pgvector) for semantic retrieval, contextual chat, and personalized recommendations. Designed modular components for authentication, monitoring, and error recovery, ensuring enterprise-scale deployment. Collaborated with product teams to translate user-facing requirements into deployable ML features and created internal tooling for rapid testing, fine-tuning, and deployment of models.

Machine Learning Engineer at Zalando

Developed and deployed computer vision services classifying millions of product images via production-ready REST/gRPC APIs, powering catalog enrichment and search improvements. Engineered real-time recommendation APIs processing live interaction streams for low-latency personalization. Integrated graph-based search and anomaly detection into fraud-prevention workflows, packaging them as reusable components. Built feature-flagged A/B testing infrastructure for ML features in production. Designed training-to-inference workflows on AWS SageMaker, exposing models as microservices and integrating them into end-user shopping experiences. Implemented time-series demand forecasting APIs to support supply chain and inventory optimization. Collaborated with frontend/backend teams to embed ML APIs into web/mobile flows. Established data versioning, feature encoding, and model validation guidelines.

AI Specialist at Capgemini

Delivered AI-enabled applications for financial, healthcare, and automotive clients by packaging ML algorithms into production-ready microservices. Built multilingual NLP services for customer-service automation, exposing chatbot APIs that integrated with enterprise CRMs and reduced manual case handling. Developed predictive maintenance APIs processing sensor IoT data in real time, enabling proactive equipment monitoring and direct integration into industrial dashboards. Migrated on-prem ML workloads to Azure ML and Azure Databricks, transforming ad-hoc scripts into standardized, reusable pipelines. Implemented ensemble forecasting services providing stable demand predictions for client planning tools.