JakubWujec
MLOps Engineer specializing in Large Language Models and production ML systems. Expert in deploying and scaling AI solutions using MLOps frameworks.
Work
MLOps Engineer
Building NPS prediction models and leveraging LLMs to enhance application performance. Working with AWS infrastructure to deploy and scale machine learning solutions for pharmaceutical applications.
Head of AI & Co-founder
Architected AI-powered technical interview platform with real-time voice processing. Built conversation engine with multi-model AI integration, WebSocket audio streaming, and intelligent assessment system reducing manual screening by 85%.
Data Scientist
Developed ETL pipelines using PySpark and BigQuery processing 100M+ records daily. Built Champion Challenger Framework for automated ML pipeline optimization and CRM models for user behavior prediction.
Machine Learning Engineer
Upgraded automatic flat valuation model with new features and optimized hyperparameters. Implemented DBSCAN clustering for geospatial anomaly detection with Dash visualization dashboard.
Data Scientist
Designed and built recommendation engine from scratch based on user choices and personal values. Created automated data pipelines for text processing from crawlers with MongoDB storage.
Education
MA Data Science
BS Computer Science and Econometrics
Technical Skills
Machine Learning
Python & Development
MLOps & Infrastructure
Cloud & Data
Projects
MockIT - AI-Powered Technical Interview Platform
AI interview platform with real-time voice processing and "Moki" AI interviewer. Built conversation engine with multi-model integration and intelligent assessment system.
KiedyKredyt - Advanced Mortgage Timing Analysis Platform
Sophisticated financial analysis platform using Monte Carlo simulations with 10,000 scenarios to provide data-driven recommendations for optimal mortgage timing.
Sandri - Enterprise Knowledge Management AI
Led development of RAG-based AI system for logistics company, integrating PDFs, FAQs, and video transcripts. Successfully delivered in 4 months.
Automated Highlights Detection
Master's thesis using YouTube's 'Most Replayed' data for highlight detection in long-form videos. Achieved competitive performance with UMT transformer architecture.
Champion Challenger Framework
Automated ML pipeline conducting monthly training to challenge champion models, enabling continuous optimization and deployment at PKO BP.
Stock Price Prediction using GAN and BERT
Bachelor's thesis combining Generative Adversarial Networks with BERT sentiment analysis to predict gaming industry stocks. Achieved profitable trading strategies for 3/4 companies tested.
Achievements
Supervision Hack 3 Winner
Led winning team in Depo Predator Challenge, developing advanced depot optimization system with 25% efficiency improvement.
Warsaw Econometric Challenge Winner
First place in Poland's premier econometric competition using innovative time series forecasting and ensemble methods.
Warsaw Econometric Challenge Winner
Breakthrough first-time victory combining traditional econometrics with machine learning techniques for superior forecasting.
ING Lion's Den - 3rd Place
Podium finish in premier fintech competition with AI-powered financial risk assessment system achieving 94% accuracy.
Let's Connect
Always interested in discussing machine learning innovations, research collaborations, and opportunities in AI development.