Dedicated to empowering organizations through data-driven and artificial intelligence solutions to optimize workflows and enhance revenue streams.
With over two decades of comprehensive experience in Computer Science across industry and academia, I specialize in devising innovative solutions that streamline operations and boost profitability, adhering to the principle of simplicity (KISS paradigm). My expertise spans the entire spectrum of the data processing pipeline – from data collection and modeling to analysis, insight extraction, and the deployment of AI solutions in production environments. I offer strategic guidance to companies embarking on AI-powered initiatives and assist them in navigating complex data transformations.
Throughout my career, I’ve led dynamic teams within startup environments, leveraging my expertise to drive innovation and optimize performance. As a trusted consultant and advisor, I’ve provided strategic guidance to companies navigating complex challenges and embarking on transformative AI initiatives. Additionally, I’ve played a pivotal role in mentoring and developing talent, guiding individuals through the intricacies of industry-specific projects and fostering their professional growth. With a knack for distilling complex concepts into actionable insights, I excel in delivering impactful presentations at industry conferences and driving the adoption of cutting-edge technologies. My commitment to excellence, coupled with a keen eye for efficiency, has consistently delivered tangible results, propelling organizations toward success in the ever-evolving landscape of data-driven solutions.
PhD in Computer Science, 2015
University of Fribourg -- Switzerland
MSc in Computer Science, 2010
University of Lodz - Poland
MSc in Computer Science, 2009
Université Claude Bernard (Lyon I) - France
BSc in Computer Science, 2006
WIT under the auspices of Polish Academy of Sciences
For a startup specializing in fitness technology, I developed an AI-powered personal fitness coach. This system leverages advanced language technologies, including GPT-4 and Command R+, to generate personalized and adaptive training plans tailored to individual user needs, setting it apart from generic, one-size-fits-all fitness solutions.
In this endeavor, I led a groundbreaking RAG project aimed at transforming document and information retrieval with provenance tracking for a valued client. Leveraging state-of-the-art technologies such as OpenAI, Neo4j, and LangChain, I orchestrated a comprehensive solution tailored to meet the unique needs of the client.
In this project, I spearheaded the development of a solution aimed at automating the summarization of customer service support conversations. The primary objective was to provide management with succinct insights into the interactions transpiring through our customer service chat platform.
In this project, I collaborated with a client developing a revolutionary non-invasive glucose monitoring system. Our initial focus was on calibrating the physical device to accurately convert raw signal measurements into glucose levels. We began by analyzing chronoamperometry data to identify optimal parameters for the device and establish a robust calibration method.
This project aimed at providing means to compute how likely it is that two bitcoin addresses belong to the same entity.
The goal of this project was to develop a Generative Adversarial Network (GAN) capable of generating unique logotypes for teams on a gaming platform. The trained GAN architecture enabled the creation of 1M unique logotypes for the players of the platform.
This project aimed at analyzing customers’ bitcoin transactions in order to provide their liquidity estimation and spending analysis.
The client was facing trouble collecting and analyzing data on customers using their API platform. I was requested to research and develop a solution to enable data analytics and visualization on their platform. The delivered solution allowed my client to understand their clients’ behavior.
The goal of this project was to research and evaluate deep reinforcement learning methods to improve order placement in limit order markets.
This project aimed at analyzing bitcoin graph data and providing sophisticated information on bitcoin transaction data and aggregate on-chain statistics derived from graph data. The outcome statistics in the form of time series were later used in predictive machine learning models.
The goal of this project was to develop a Deep Learning powered solution detecting a disease on medical X-ray images. The model I have developed achieved an AUC of 0.98. The solution is part of a bigger system which is developed to assist medical doctors in their daily work.
The goal of this project was to develop a machine learning backed investment strategy based on multiple binary positive and negative signals sourced from price, market, and environmental analysis. The strategy I have developed performed 4x better than the baseline.
The goal of this project was to build an ETL system processing live online bitcoin transactional data and loading it to Neo4J graph database on the fly. The data included not only block chains but all details about transactions and interactions between addresses.
The goal of this project was to model blockchain historical data as a graph and to build an ETL system loading the entire data (blocks, transactions, inputs, outputs, addresses) into a Neo4j graph database. The solution I designed performed 10x faster than the best-reported system.
The goal of this project was to develop an AI backend engine for an intelligent decision support system which asses an ischemic stroke risk. The system was developed in collaboration with a health insurer to allow preventive interventions for patients with high risk of a stroke.
The goal of this project was to build an ETL system collecting and computing data about trades and trading positions for various cryptocurrencies. The data about trades and trading positions were integrated on the fly from multiple trading platforms via REST APIs and web socket.
The goal of this project was to develop a rules-based decision support system. The solution was assisting medical doctors to make a fast decision about blood transfusion. The project was conducted in collaboration with medical doctors as domain experts.
The goal of this project was to develop an AI-backed decision support system helping medical doctors to decide about blood transfusion. In collaboration with multiple hospitals, we collected historical data about blood transfusions, patients, and medical tests relevant to blood transfusion.
The goal of this project was to build an ETL system collecting, normalizing, and aggregating data about the popularity index (search volume) of various keywords from Google Trends. The system had two subsystems one for pulling historical data in bulk, the other for current data.
SolarData is a platform allowing to simulate the amount of energy produced by a photovoltaic energy system in a specified geographic location weather data and parameters of the photovoltaic system like number and type of solar panels, the angle of the installation, etc.
The goal of this project was to develop an AI backend for a frequently asked questions chatbot. A client had collected a significant amount of questions and answers that were used to train machine learning natural language processing models.
The goal of this project was to build an ETL system collecting and aggregating data about various crypto assets. The data about blocks and transactions on the crypto assets was fetched and integrated on-the-fly from multiple data sources via REST APIs.
The goal of this project was to develop an anomaly detection pipeline for business intelligence systems in order to trigger a BI report distribution - via slack or e-mail - when an anomalous situation was detected or send an alert to predefined groups of managers.
TripleProv is an in-memory RDF database capable to store, trace, and query provenance information in processing Linked Data queries. TripleProv returns an understandable description, at multiple levels of granularity, of the way, the results of a SPARQL query were derived.
dipLODocusRDF is a system for Linked Data data processing supporting both simple transactional queries and complex analytics efficiently. dipLODocusRDF is based on a hybrid storage model. DiploCloud is distributed version of dipLODocusRDF.