Piotr Pawłowski Email and Phone Number

- Developed an explainable heart pathology classifier with Deep Learning and Computer Vision methods for one of the largest Polish medical facilities. The development included conducting extensive anonymization and preprocessing of X-Ray images using OCR and Named Entity Recognition. For this purpose, utilized Azure ML, PyTorch and image processing tools.- Designed and developed a conversation steering and information gathering, LLM-powered chatbot for compliance assessment against European regulation in cybersecurity space- Led the team that successfully designed and developed LLM-powered assistants for RAG (multiple knowledge bases), translation, entity extraction and summarization tasks in legal, tax and assurance domains. For this purpose, utilized Azure, langchain, llama-index, different vectorstores and databases.- Implemented and integrated an efficient and transparent pricing optimization algorithm for financial products in a large Polish bank. Collaborated with the client's MLOps team to enhance their in-house ML Platform based on Google Cloud Platform and Open Source Software. Implemented ML pipelines in Kedro / Airflow, then deployed and monitored models with Seldon.- Conducted technical interviews for Senior Data Scientists- Participated in sales effort by engaging in activities to attract and secure new clients.

- Led the team of Data Scientists that successfully designed, built and integrated an intelligent logs analysis tool into an existing software solution used by multiple clients. For this purpose, utilized Team Data Science Process lifecycle, text and process mining tools, Deep Learning algorithms for Sequence Classification, Sequence Tagging and Kedro for ML pipelines. Conducted a controlled experiment in order to prove the value over an existing approach.- Developed and enhanced an internal MLOps Platform. For this purpose, utilized Docker, Kubernetes, several MLOps frameworks and other cloud native solutions. Built tools and APIs to enable users with varying amounts of technical expertise.

- Designed and implemented ML pipelines in Kubeflow and PyTorch that extract contents and anonymize large corpora of legal documents in a variety of formats (docx, pdf). During the process, utilized state of the art Deep Learning algorithms for Sequence Tagging, Language Detection and other supporting NLP tasks.

- Designed and built search engines (ELK stack) for multiple projects centered around different types of documents (legal documents, news articles, marketing materials)- Developed ETL pipelines with Apache Airflow and Apache Spark for large scale document processing.

- Collaborated with business and domain experts (tax, law, finance, insurance) to gather and share insights about data-centered processes. Built interactive prototypes and visualizations (Flask, FastAPI, plotly, matplotlib, streamlit)

- Consolidating and analyzing data collected via Google My Business;- Analyzing competitors' actions with Google AdWords and Ahrefs;- Creating various, interactive reports in MS Excel and Google Spreadsheet;- Analyzing web traffic on company's websites and verifying deviations with Google Analytics;