Wojciech Krukar

PROFESSIONAL EXPERIENCE

Lead Data Scientist Mastercard | March 2022 – Oct 2025

• Architected and deployed distributed ML pipelines on AWS supporting real-time fraud detection

processing 143+ billion transactions annually through Decision Intelligence Pro system, achieving

20-300% improvement in fraud detection rates and 85% reduction in false positives through

advanced algorithm development and feature engineering

• Implemented Hugging Face Transformers for generative AI fraud detection scanning 1 trillion data

points in under 50 milliseconds, reducing manual investigation time by 60% while maintaining

regulatory compliance.

• Fine-tuned large language models (LLMs) on transaction narratives using advanced natural language

processing and statistical analysis, improving conversation summary accuracy by 45% and enabling

data storytelling for executive stakeholders.

• Led cross-functional team of 12+ data scientists in designing, building, and testing computer vision

engineering solutions for merchant verification, contributing to $118 billion annual fraud prevention

savings through advanced feature engineering and transfer learning techniques.

• Optimized reinforcement learning algorithms in cloud computing environments, reducing system

latency by 25% and improving recommendation accuracy by 18% through sophisticated algorithm

development and hyperparameter tuning.

• Designed A/B testing processes and analytics dashboards using statistical software and hypothesis

testing methodologies, driving data-informed business decisions and enhancing marketing targeting

effectiveness.

• Developed data architecture supporting big data analytics infrastructure with comprehensive data

security measures, ensuring GDPR compliance and data transformation processes for global payment

systems.

• Integrated MLOps pipelines with Kubeflow and Docker, reducing operational costs through

automated data lifecycle management and implementing data profiling techniques for quality

assurance.

• Applied time series analysis and predictive modeling techniques for transaction forecasting

achieving 95% accuracy in fraud pattern prediction and enabling proactive threat mitigation

strategies.

• Built clinical NLP pipelines for medical chart processing: sectioning, de-identification, entity

detection (problems, meds, procedures), and ICD-10/CPT suggestion with human-in-the-loop review,

improving coder throughput by 32% and reducing DRG upcoding risk alerts by 18%.

• Deployed PySpark-based RCM anomaly detection (payment accuracy/smart reconciliation) over

1B+ claim lines, cutting denial write-offs by 6.4% and accelerating cash posting by 19 hours on

average.work.

• Created medical image triage models (X-ray/ultrasound) with MONAI and U-Net variants;

prioritized review queues and reduced average turnaround time by 21%.

Senior Data Scientist General Motors | June 2021 – Dec 2021

• Delivered predictive maintenance models using time series analysis and statistical analysis,

increasing anomaly detection accuracy by 20% across 1,200+ daily vehicle production lines through

multi-billion NVIDIA partnership.

• Implemented computer vision engineering solutions for quality inspection automation, reducing

processing time by 30% and achieving $7.4M annual cost savings through advanced algorithm

development.

• Deployed reinforcement learning optimization for digital twin manufacturing simulations using

NVIDIA Omniverse platform, improving manufacturing efficiency by 17% with zero safety

incidents.

• Designed data architecture for connected vehicle platforms supporting big data analytics on 25K+

sensors, enabling predictive analytics for autonomous driving systems development.

• Applied feature engineering techniques to battery management systems data, enhancing electric

vehicle performance optimization by 15% through sophisticated data transformation processes.

• Integrated cloud computing resources (AWS, Azure) for scalable data processing of automotive IoT

data, implementing comprehensive data security frameworks for connected vehicle platforms.

• Utilized statistical software and data visualization tools to create executive dashboards, enabling data

storytelling for C-suite decision-making regarding electric vehicle expansion to 12 models by 2026

Lead Data Scientist Hypergiant | January 2019 – December 2020

• Designed patented semantic segmentation models for defense and space applications, securing $61

million U.S. Air Force JERIC2O contract for Joint All-Domain Command and Control systems

• Implemented reinforcement learning algorithms for autonomous satellite operations, achieving 35%

mission efficiency improvement and 100% mission success rate for NASA, USAF, and Space Force

clients.

• Developed computer vision engineering solutions for satellite imagery analysis using advanced CNN

architectures, accelerating target identification workflows by 35% through sophisticated feature

engineering.

• Implemented feature engineering techniques for defense sensor data analysis supporting $950

million Department of Defense IDIQ contract awards and multi-level security clearance systems.

• Integrated AI into astronaut equipment and OCR tools using natural language processing and

statistical analysis, enhancing safety protocols and operational efficiency for space missions.

• Built scalable AI prototypes using Kubeflow and distributed computing technologies on cloud

platforms, implementing comprehensive data security measures meeting DoD STIG requirements.

• Applied time series analysis and predictive modeling to satellite constellation management, reducing

ground-station handover latency by 28% through advanced algorithm development.

AI/DL Data Scientist

Dell EMC | November 2017 – December 2018

• Led ML system design on big data infrastructure using Ready Solutions for AI, achieving 2.9x

performance improvement over competitors and reducing AI deployment time by 6-12 months for

Fortune 500 enterprises.

• Implemented automated 5-click data science workspace provisioning using Jupyter notebooks and

statistical software, boosting data scientist productivity by 30% through streamlined data lifecycle

management.

• Optimized deep learning model training performance using mixed-precision techniques and

advanced algorithm development, improving GPU utilization from 68% to 92% and reducing energy

costs by 25%.

• Designed data architecture for enterprise AI democratization supporting cloud computing platforms

(AWS, Azure), enabling comprehensive data transformation and feature engineering workflows.

• Applied statistical analysis and hypothesis testing methodologies to validate model performance,

ensuring data quality assurance and implementing data profiling techniques for enterprise datasets.

• Integrated MLOps pipelines with comprehensive data security frameworks, enabling scalable

deployment of machine learning models across diverse enterprise environments while maintaining

data governance standards.

Data Scientist/Co-founder

ITSolidServices LLC | February 2014 – October 2017

• Developed AI-powered medical app 'Bluepad' using computer vision engineering and machine

learning, achieving 95% diagnostic accuracy on mobile devices with HIPAA-compliant data

architecture.

• Designed comprehensive data security and transformation infrastructure for healthcare clients,

implementing advanced encryption and data governance frameworks passing zero critical audit

findings.

• Applied statistical analysis and feature engineering to clinical datasets, identifying key patterns and

opportunities for predictive healthcare analytics and personalized treatment recommendations.

• Delivered AI training workshops for 120+ medical professionals, increasing healthcare AI adoption

rates by 45% through effective data storytelling and technical documentation.

• Implemented natural language processing solutions for electronic health records, reducing provider

documentation time by 25% through automated data annotation and algorithm development. 
 

PhD in Computer Science