Michał Kordyzon - personal site

(info about my work)

—————————————————————————————————————————————————————–

Hi there 👋

My new technical blog: mlog

I am a data scientist working @IBM

genAI related projects

Starting in end of 2022 the era of genAI started. Since then, I’ve been working mainly on genAI projects using open-source and proprietary watsonx.ai (IBM’s platform for LLMs). These projects focused on RAG applications, automated report generation (from both tabular data and documents), and “intelligent” search.

2025 is the year AI agents. For me it’s important also for personal reasons. I’ve been fascinated by agentic AI since 2017. As an example I had a session about Reinforcement Learning Agents in 2020 on Think Warsaw where I was discussing with IBM Research guest the ethicacl agent created by IBM (the agent playing Pacman ‘ethically’, and avoided ‘killing’ ghosts :) :) :) . Here is the link: ethical Pacman.

ML related projects

• Machine Learning & Data Governance

  I have engaged in many projects as an ML Engineer, or as a Data Scientist. See the notes about one interesting ML project as example. I have used theere a tweedie regressor with transformers to count pure premiums for insurance product. Model and transformer were hosted on Watson Machine Learning and they used Watson OpenScale monitors to scan bias, explainability, fairness and drift.

• Natural Language Processing

  I have built Miniman Viable Products using NLP, as an exaample, one case was about social listening: scrapping & analysis of social content using python libraries and translating the insights into business value.

• Optimization

  See my two Decision Optimization workshops, I’ve made them for clients starting to use IBM Decision Optimization (aka prescriptive analytics). It’s a great way to enrich machine learning projects, and funny thing is that so few client even know about it. For optimization projects I am using CPLEX, and DOcplex, two IBM libraries that can help build our dynamic programming models efficiently.

• Visual Recognition

  Using Detectron, Yolo and FAster R-CNN I delivered value in many visual recognition projects. The platform for this was Maximo Visual Inspection. As example, I was creating a model for one client where AGV carts were inspected for security by the model, which saved time on and gave high accuracy comparing to human inspection. The movie made after project completion is here.

• Neuro-symbolic AI

  It’s always beneficial to attend a good summer school like this one. Neural-symbolic AI is kind of deductive system allowing you to put a reasoning layer on top of your neural nets. You will find general info about neural-symbolic AI here. If you want the hands on experience, take all activities needed to get this badge.

• Qiskit

  In 2020 I have spent a summer (and few consecutive months) learning about Qiskit, great Python SDK for quantum programming. I have contributed to Qiskit in many ways, attended hackathons, and enjoyed being the part of the community. I am glad IBM is making such risky bets. Google’s Larry Page famously said: “Especially in technology, we need revolutionary change, not incremental change.”

• IBM Z

  As an engineer I have grew profesionally on IBM Z platform, the most stable, reliable, and long-lived server platform on the planet. IBM Z is unique and very attractive for both technical folks and the bussiness, it’s like Warren Buffett of engineering. IBM Z prefers: reliability over hype, investment continuity over price wars and technology unexpected ‘End of Marketing’, quality over ‘nice features’. My last project for IBM Z was in 2018, when I had a priviledge to co-author z14 servers IBM Redbooks documentation (see links below).

Check some of my publications:

• article about Qiskit
• article about Decision Optimization
• System z IBM Redbook that I have co-authored
• System z IBM Redbook that I have co-authored