Senior Data Scientist & ML Engineer with 5+ years of experience
End-to-end machine learning solutions — from prototyping to production deployment.
Autonomous systems that reason, plan, and take action. From conversational chatbots to complex multi-step workflows with tool use, memory, and retrieval-augmented generation.
Text understanding, generation, and transformation at scale. Sentiment analysis, document processing, custom LLM fine-tuning, and end-to-end speech-to-text pipelines.
Deep learning architectures for vision, segmentation, and classification. CNNs, autoencoders, and custom architectures — trained, optimized, and deployed for production.
Real-time monitoring and fault detection for industrial systems. Autoencoder-based methods on high-frequency sensor data, with edge deployment down to microcontrollers.
Data-driven optimization for marketing campaigns, logistics, and recommender systems. Clustering, mathematical programming, and algorithmic decision support.
Multi-horizon forecasting for sales, demand, and operational metrics. Deep learning and statistical models for accurate, actionable predictions with confidence intervals.
RAG pipelines, vector databases, and intelligent document search. Connect LLMs to your proprietary data for grounded, accurate, and citation-backed responses.
A selection of production ML systems and professional experience across industries.
Building an AI startup for your social media inbox.
Led multiple data science and ML projects across diverse clients. Specialized in anomaly detection, NLP, OCR, and LLM integration for enterprise solutions.
Developed numerous data science and ML projects focused on data engineering, pipeline architecture, and customer analytics.
Software development on IBM Data Gate — enterprise data integration and replication platform for hybrid cloud environments.
Researched ML techniques for Master Data Management. Developed a power ranking system for IBM Watson's partnership with the Overwatch League esports.
Conducted research on radar sensing techniques for autonomous driving and advanced driver-assistance systems.
Implemented autoencoder-based anomaly detection on high-frequency sensor data for quality assurance in automotive production lines.
Developed and optimized two OCR pipelines for automated extraction and validation of complex vehicle documents including security verification checks.
Built a custom dataset and trained a segmentation network to detect structural defects, enabling automated quality inspection on production lines.
Integrated retrieval-augmented generation workflows with local LLMs, vector databases, and LangChain for intelligent data linking and dynamic code execution.
Designed feature-based anomaly detection for industrial press systems and migrated a deep learning model from TensorFlow to PyTorch.
Deployed real-time infrared-based monitoring of industrial process interfaces with AI-driven defect detection and cycle-based analysis.
Created a thermal anomaly detection model ported from Python to C for embedded deployment on microcontrollers with constrained resources.
Designed a scalable data pipeline with clustering and record linkage to analyze customer behavior. Integrated external data via APIs and databases.
Built a deep learning model for sales forecasting, improving accuracy through multivariate time series analysis.
Developed AI-based optimization for marketing campaigns using clustering and mathematical optimization methods.
A collection of personal projects — from games to data visualizations — built for fun and curiosity.
A timed flag identification quiz game with leaderboards, streak multipliers, and customizable key bindings. How many flags can you identify in 30 seconds?
A statistical deep-dive into how much of sports outcomes are driven by luck versus skill — with interactive visualizations and analysis across multiple leagues.
Self-driving Cars Lecture Competition
2024German Physical Society
2014US 12,138,554 B2
2021Interested in working together? Reach out and let's discuss your next AI project.