AI LEADERSHIP · SOFTWARE ARCHITECTURE

Marcin Stasko.

Technology Leader | AI & Software Architect | Founder & CEO

Wrocław, Poland · English C1 · Polish native

I build AI systems and the teams that deliver them.

I lead technology teams and still work hands-on with AI architecture. At LG Energy Solution, I built digital transformation capability and coordinated work across more than 30 engineers. Today, I run Science For People, combining AI consulting, implementation, education and product development.

Technology leadership · AI architecture · Advisory

Download CV View experience

PDF · 3 pages · English

mstasko@s4p.eu +48 663 222 322 LinkedIn ↗

30+engineers across delivery teams

50+consulting projects · 16 countries

3,000+consulting delivery hours

14,000+learners and professionals reached

01 / Projects

Selected AI projects.
From data to production.

Selected work at LG Energy Solution, from factory equipment to enterprise knowledge. Explore the scope, architecture and implementation behind each result.

LG ENERGY SOLUTION / 2021–2026

60production machines

10 signal types

Each square represents one of the 60 machines in scope.

Predictive maintenance

An end-to-end program connecting industrial data research, signal processing and feature engineering with trained models, application integration and deployment.

Reported project savings: >$10M annually

PythonSignal processingModel trainingMLOps

FROM INPUT TO OPERATION

Machine signals
Denoising & features
Model training
Application
Deployment

Savings are reported for individual projects and should not be added together.

02 / Experience

Professional experience.

My consulting practice began in 2017 and continued alongside my roles at LG.

2017–PresentCURRENT PRACTICE01

Independent Consultant; subsequently Founder & CEO

Science For People / S4P OÜRemote / International

Completed more than 50 consulting projects in 16 countries, including over 3,000 consulting hours. The work covered AI strategy, architecture reviews, technical development, integration and implementation planning.
I advise founders and leadership teams on use cases, investment priorities, governance and practical ways of working with AI. I turn business goals into roadmaps with clear ownership.
I develop the Science For People platform for consulting, implementation support and education. Its AI tools connect prompt specifications, evaluation, human review and reusable knowledge with a clear record of decisions and completed work.
Lead Try-Doo, an Android/web AI hairstyle assistant combining facial landmarks, photo analysis, recommendations, generated previews, preferences and salon-ready instruction cards.
I also look after the product's day-to-day operation, from credits and Google Play Billing to content reports, data deletion, rewarded ads and services running on GCP Cloud Run.
I help organizations move LLM and RAG solutions from proof of concept to production. My workshops, courses and mentoring have reached more than 14,000 learners and professionals.

2021–March 202602

AI/Digital Transformation Manager

LG Energy SolutionBiskupice Podgórne, Poland

Built digital transformation capability and coordinated work across more than 30 engineers in software, data science, automation and manufacturing. Managed technology budgets, server procurement, KPIs and architecture under strict on-premises security requirements.
Led predictive maintenance across 60 machines and 10 signal types, from signal processing and model training to deployment; reported project savings exceeded $10M annually.
Led computer vision for seven defect-detection problems across 50 production lines (five supervised, two unsupervised); reported project savings exceeded $45M annually. Used Python, PyTorch and OpenCV.
Directed company-wide autocalibration from equipment and PLC integration through data processing, auto-optimizing algorithms, software deployment and machine operating procedures; developed battery formation and electrochemical quality analytics.
Architected and deployed an internal LLM/RAG platform with Milvus and Django. Developed agentic applications, document processing and transcription workflows; conducted separate LoRA adaptation research using enterprise knowledge.
Led SAS-to-TIBCO Spotfire migration, data-lake/staging integration and tailored BI. Introduced modular Docker/MongoDB services, CI/CD, MLOps and YouTrack API delivery analytics.
Represented LG through international keynotes, lectures and media work. Supported university curricula, an electromobility laboratory and chemical/IT expertise for complex legal and procedural matters.

2020–202103

Senior Data Scientist

LG Chem / LG Energy SolutionBiskupice Podgórne, Poland

Led a Six Sigma Black Belt project delivering $28M in savings; magnetic-separator diagnostics and algorithm upgrades delivered $2M annually. Automated SAS/Python reporting, saving 25,000+ work hours annually.
Built APAS manufacturing BI, presented it to the LG Chem CEO and supported its worldwide distribution within LG. Developed data-warehouse-backed End-Use/IPR customs traceability and interprocess ML.

2018–202004

Senior Defect Engineer

LG ChemBiskupice Podgórne, Poland

Led a Six Sigma Green Belt project generating $2.3M annually; investigated defects across electrodes, assembly and formation, and researched chemical treatment, separation and industrial cleaning.

03 / Skills

Technical expertise.

From model development and software architecture to industrial integration and production operations.

AI, machine learning & signals

Python · PyTorch · Scikit-learn · XGBoost · Pandas · model training · feature engineering · forecasting · time-series analysis · denoising · spectral analysis · anomaly detection

LLMs & knowledge systems

Large language models (LLMs) · retrieval-augmented generation (RAG) · LangChain · LangGraph · Langfuse · Milvus · vector search · agentic workflows · LoRA adaptation research

Computer vision & industrial integration

OpenCV · supervised defect detection · unsupervised image anomaly detection · camera integration · facial landmarks · PLC integration · autocalibration · battery formation analytics

Software & architecture

Python · SQL · Django · REST APIs · modular services · microservices · Git · YouTrack API · application integration · system architecture

Data platforms & analytics

PostgreSQL · MongoDB · Redis · Milvus · SAS · TIBCO Spotfire · data warehouses · data lakes · staging zones · business intelligence · manufacturing traceability

Cloud, infrastructure & delivery

Google Cloud Platform · Cloud Run · Docker · Kubernetes · on-premises deployment · Apache Airflow / DAGs · CI/CD · MLOps · server procurement · production monitoring

How I lead
and work with people.

Cross-functional leadership

Built delivery capability across data scientists, developers, automation engineers, technicians and manufacturing specialists.

Executive communication

Translate technical choices, investment needs and implementation trade-offs for founders and senior stakeholders.

Critical thinking & evidence

Combine scientific research, root-cause analysis and expert testimony; explain uncertainty without overstating conclusions.

Prioritization & accountability

Connect roadmaps, budgets, resource constraints and KPIs to delivery responsibilities and operational outcomes.

Mentoring & knowledge sharing

Design workshops, courses and mentoring for engineers, analysts, managers and people adopting AI.

Collaboration under pressure

Bridge business and engineering perspectives through structured communication, conflict management and practical problem-solving.

Languages

PolishNative

Native communication

EnglishC1

Advanced professional proficiency

LatinB1

Intermediate

04 / Career story

From scientific research
to technology leadership.

My work connects the physical world with data, software and business decisions. Scientific research taught me to examine evidence; manufacturing taught me to make it useful. AI architecture and leadership brought these disciplines together.

Explore my career story

012010–2017

Understand the mechanism.

Chemistry, engineering and international research built the habit of testing assumptions against evidence. Work spanned combustion emissions, biocatalysis, fluorescent probes and protein labeling.

SCOPE & EXPERIENCE5 research institutions

Opole University · Opole University of Technology · Graz · Karolinska · Polish Academy of Sciences

ResearchStatistical reasoningExperimental methods

I have consulted since 2017, alongside my roles at LG.

05 / Education & research

Education & qualifications.

View credentials

Opole University

Bachelor’s & Master’s degrees

Research period: 2014–2017. Chemistry, biotransformation and enzymatic processes.

Opole University of Technology

Engineer

Study period: 2010–2014. Individual Study Programme; thesis on polycyclic aromatic hydrocarbons in biomass-combustion exhaust gases.

Professional qualifications

Leadership & delivery

ITIL Master
PRINCE2 Agile Expert
Professional Scrum Master II
Lean Six Sigma Black Belt — LG Chem
Lean Six Sigma Green Belt — LG Chem

AI & data

Probabilistic Graphical Models — Stanford University
Advanced Data Science — IBM
SAS Certified Specialist: Machine Learning Using SAS Viya 3.5
SAS Certified Predictive Modeler Using SAS Enterprise Miner 14
DAG Authoring for Apache Airflow

Additional professional study

ITIL: Foundation; Strategic Leader; Direct, Plan and Improve; Digital and IT Strategy; High-velocity IT
Forensic Science and Criminal Justice — University of Leicester
Food Safety / Quality Systems — LUQAM & Bureau Veritas: HACCP, ISO 22000, IFS, BRC

Research appointments

2014–2017

Research

Opole University · Opole, Poland

Studied microbial steroid-hormone biotransformation and oxidative enzyme activity. As president of the Chemical Association, organized scientific events and workshops and represented students on the faculty board.

2014–2017

Research

Opole University of Technology · Opole, Poland

Conducted chromatographic analysis of combustion emissions and developed a catalyst supporting biofuel combustion.

June 2016 – October 2016 · 5 months

Visiting Researcher

Karolinska Institutet · Greater Stockholm Metropolitan Area, Sweden

Completed a research placement in Medical Biochemistry and Biophysics, focused on modifying a bacterial expression vector and selectively labeling selenoproteins with fluorescent markers.

June 2015 – October 2015 · 5 months

Research Intern

University of Graz · Graz, Austria

Completed a biocatalysis research placement at the Institute of Organic and Bioorganic Chemistry, investigating enzymatic phosphorylation of alcohols in low-water solvents. The research later contributed to a ChemSusChem publication.

June 2014 – October 2014 · 5 months

Research Intern

Institute of Organic Chemistry, Polish Academy of Sciences · Warsaw, Poland

Conducted research on the synthesis of fluorescent probes for enzymatic assays.

Publication, writing & speaking

Peer-reviewed publication

Tasnádi, G.; Staśko, M.; Ditrich, K.; Hall, M.; Faber, K. “Preparative-Scale Enzymatic Synthesis of rac-Glycerol-1-phosphate from Crude Glycerol Using Acid Phosphatases and Phosphate.” ChemSusChem, 2020, 13(7), 1759–1763.

DOI: 10.1002/cssc.201903236 ↗

Writing, speaking & education

Technical author on predictive maintenance and AI; international keynote speaker; university curriculum and electromobility laboratory contributor. PeopleCert community and Polish ITIL Version 5 beta examination participant.

My technical writing covers data acquisition, denoising, stationarity, time-domain and spectral features, and anomaly detection for predictive maintenance, alongside AI implementation, data science and manufacturing transformation.

Read technical articles