Work Experience

Work Experience — Konrad Chmielinski
Konrad Chmielinski · Career Overview

5+ years building enterprise automation and AI solutions across global manufacturing, banking, and steel industry environments.

RPA · AI · Data Engineering 5+ years experience 2 employers · 10+ countries Open to remote · Full-time / B2B

Locations of clients and organisations I have cooperated with, delivered solutions for, and supported across their business processes:

My career has followed a clear trajectory: from building reliable, production-grade RPA automation at a major international bank, to expanding into AI-enhanced workflows and data pipelines at one of the world’s largest steel producers. Each role has added a new layer — deeper technical complexity, broader geographic scope, and a growing focus on applying AI where it creates real business value.

5+
years of professional experience
30+
bots & automated processes delivered
12
significant projects across 2 employers
10+
countries served globally
Internationale Nederlanden Groep Hubs Poland B.V. Sp. z o.o.
Process Automation Developer — Intelligent Automation & Business Analytics (GKYC)
November 2021 — October 2023 2 years

ING is one of Europe’s largest banking groups. The Poland hub acts as an international delivery centre for technology and operations, serving ING entities across Western Europe. I worked within the Global Know Your Customer (GKYC) team — a compliance-critical function responsible for customer due diligence, identity verification, and financial risk assessment.

Working in a compliance environment meant that accuracy, auditability, and regulatory alignment were non-negotiable. Every automation had to be documented, tested against real regulatory requirements, and designed to flag exceptions clearly rather than silently processing bad data. It was in this role that I built the foundations of how I approach production automation today.

Beyond technical delivery, I was also actively involved in promoting automation adoption internally — running training sessions on SAIO (ING’s proprietary RPA platform) and helping colleagues understand how to scope and build their own automation workflows.

The headquarters building of ING Hubs Poland located in Katowice, Poland.

70%
processing accuracy on multi-market document pipeline
6+
Western and Central European markets: LU, NL, FR, BE, PL, SK
2yr
experience in compliance-critical automation
Risk Assessment Classifier ML · CDD
Built, enhanced, and adapted a classifier to estimate organisational risk for financial security, fraud, misappropriation, and identification of harmful or non-compliant entities. Updated the system iteratively as business requirements and regulatory frameworks evolved — improving accuracy and supporting proactive risk management for Customer Due Diligence (CDD) across ING’s Western European operations.
Document Scan Intelligence — Western Europe NL · FR · BE · DE
Led development of an automated pipeline for extracting and validating information from power of attorney documents and identity documents. Covered multiple document formats across four markets, each with different layouts, languages, and legal requirements. Included OCR extraction, document classification, signature presence detection, and Pydantic-validated field output. Achieved 70% end-to-end processing accuracy, substantially reducing manual verification effort and accelerating delivery timelines.
Portal Data Exchange Bots 2 bots
Built and maintained two RPA bots facilitating secure information exchange, automated file downloads, and report retrieval from multiple external portals. Selenium-based web automation with robust session management, retry logic, and structured logging. Optimised for reliability across long unattended runs.
SAIO GUI Bot & Platform Promotion Internal platform
Designed and implemented a GUI-based bot on SAIO — ING’s proprietary RPA platform — to streamline internal processes. Actively promoted SAIO adoption within the team by preparing and delivering training sessions and best-practice documentation, helping colleagues scope and build their own automation workflows independently.
Python SAIO (ING proprietary RPA) Selenium ABBYY FlexiCapture ABBYY Vantage Azure OCR Scikit-Learn OpenCV Pillow Pydantic Pandas REST APIs
ArcelorMittal Business Center of Excellence Poland
RPA Developer — Advanced Technology Center (BCoE)
October 2023 — present Currently employed

ArcelorMittal is the world’s second-largest steel producer, operating across 60+ countries. The Business Center of Excellence in Poland acts as an internal shared-services hub delivering automation, analytics, and technology projects for business units globally — from Luxembourg and India to Poland and beyond.

In this role I design, develop, and maintain end-to-end automation solutions for global stakeholders. Projects span financial data consolidation, HR process automation, compliance reporting, document intelligence, and AI-enhanced workflows. Every delivery requires coordinating with international teams, understanding business processes I have never seen before, and building solutions robust enough to run unattended across time zones.

10+
countries served by my bots
7
significant projects delivered
15+
bots delivered and maintained
60%
accuracy on AI offer generation pipeline
5yr
invoice data analysed for duplicate detection
Global Market Trend & Investment Forecasting Global scope
Designed and deployed a bot consolidating Excel data from all global ArcelorMittal business units into a centralised database, enabling multi-year market trend prediction and investment profitability analysis. Manages end-to-end data ingestion, supplemental report generation, and web application integration. Used daily by global finance stakeholders.
Chemical Analysis & Research Report Extraction Azure AI
Maintaining an Azure-based pipeline that extracts and consolidates data from diverse chemical research reports using Azure Document Intelligence. Oversees document classification, OCR workflows, data quality monitoring, and integration with centralised databases for reporting and analytics.
Global Invoice Duplicate Detection ArcelorMittal India
Developed an automated solution analysing five years of transactional invoice data to identify potential duplicates and prevent future approvals. Reduces financial risk and improves operational efficiency across global procurement workflows. Idempotent design ensures safe re-runs without creating false duplicate flags.
Global HRS Document Automation 4 bots delivered
Delivered four bots for the global Human Resources Services team, automating contract generation and data entry from financial and entity-related documents into SAP. Significantly improved processing speed and accuracy for a team serving employees across multiple countries and legal entities.
Global Compliance Reporting Automation ArcelorMittal Luxembourg
Built a bot processing reports and multimedia files to update portal data and generate formal compliance reports on workplace incidents — producing detailed unit-level statistics for regulatory authorities. Combines web portal automation, file processing, and automated PDF report generation in a single end-to-end workflow.
VBA to Python Bot Migration Modernisation
Migrated two legacy VBA bots to Python to comply with new company standards. Required reverse-engineering undocumented business logic embedded in years of macro code, rewriting in maintainable Python, and integrating with modern CI/CD pipelines — while preserving identical output for downstream stakeholders.
ArcelorMittal Park in Sosnowiec

ArcelorMittal Park in Sosnowiec, Poland.

Python Automation Anywhere A360 Azure OpenAI Azure Document Intelligence LangChain SAP GUI Scripting Pandas Jenkins CI/CD Git REST APIs SQL Pydantic
ArcelorMittal Luxembourg Downstream Solutions S.A.
RPA and AI Developer — Specialist (Secondment)
March 2024 — July 2024 Secondment · Full-time · Remote

A four-month full-time secondment to ArcelorMittal’s Luxembourg downstream operations, working remotely as an embedded AI developer within the local sales and operations team. This was a direct result of ArcelorMittal Luxembourg identifying a use case where AI could significantly reduce manual workload — and requesting a specialist from the BCoE to deliver it.

The project required understanding a highly domain-specific sales process, working with historical order data spanning multiple years and product lines, and building a production pipeline that the sales team could rely on daily — without any RPA or AI background on their side.

AI Client Offer Generation LLM pipeline
Developed an AI-driven solution using Azure OpenAI (GPT-4o) to generate personalised client proposals based on historical orders and new client inquiries. The pipeline retrieves relevant order history via REST API, constructs a structured prompt, runs LLM inference, validates the output with Pydantic, and automatically dispatches a personalised email response. Achieved approximately 60% accuracy end-to-end, significantly reducing the workload of the sales team and accelerating response times.
Azure OpenAI (GPT-4o) LangChain Prompt engineering REST API Pydantic Python Microsoft Graph API

What connects both roles

Across both employers the underlying challenge has been the same: complex, high-stakes business processes that need to run reliably without constant human supervision. At ING that meant compliance-critical identity verification across four countries. At ArcelorMittal it means financial forecasting, HR operations, and AI-driven sales workflows serving teams across three continents.

In both environments I have had to earn the trust of business stakeholders who are not technical — by delivering on time, documenting everything, and building exception handling that flags problems clearly instead of hiding them. That discipline, more than any specific technology, is what I carry from role to role.

About the Author

Leave a Reply

Your email address will not be published. Required fields are marked *

You may also like these