World experts in AI, developing cutting-edge technology purpose-built to address the most pressing challenges in enterprise data processing.
Pathway is created by people who worked in the best AI labs in the world, ex-Googlers, world's top competitive programmers, experts in Machine Learning and Complexity Science. Discover our story.
Zuzanna is the author of state-of-the-art model for forecasting of maritime trade published by the National Academy of Sciences of the USA. She holds a Master’s degree in Economics and Public Policy from Sciences Po, Ecole Polytechnique, and ENSAE, as well as a PhD in Complexity Science.linkedin
Jan is a reference figure in AI and NLP who worked at Microsoft Research, Google Brain and the MILA AI Institute. He co-authored papers with Yoshua Bengio and Geoff Hinton, two of the three Godfathers of AI. Jan is highly recognized for his research in machine learning with over 8000 citations on Google Scholar. He has a deep love for good engineering, holds PhD in neural networks, and has a background in Electrical Engineering.linkedin
Adrian obtained his PhD in Computer Science at the age of 20. He has 15+ years of experience in research. Adrian co-authored over 100 research publications, including articles in the world's best conferences & journals in different disciplines (Theoretical CS, Physics, Biology). He worked on transportation optimization, route planning, and logistics projects in collaboration with industrial partners. As co-founder of Spoj.com, the first platform and community for online training of programming, Adrian had a chance to impact the lives of millions of programmers around the globe. He is also a distinguished competitive programmer and coach. As Head of Business Development at Spoj.com, he acquired the first major contracts.linkedin
Bartek is a bronze medalist from the "Olympic Games of programming" (ACM ICPC World Finals) and published numerous research papers in Theoretical Computer Science at the world's best conferences. His expertise in data structures and randomized dynamics was key to cracking of some of the biggest efficiency challenges in Pathway™. Bartek previously worked at Google on the search engine optimization.linkedin
Former Global CEO for DHL Global Forwarding and Freight, and Management Board Member of Deutsche Post.
Business Angel, Open AI
Deep learning researcher at OpenAI. Previously part of the Google Brain team. Works on the fundamental aspects of deep learning and Natural Language Processing (NLP). Co-invented Transformers and other neural sequence models. Co-authored the TensorFlow system and the Tensor2Tensor and Trax libraries.
When Zuzanna was finishing her PhD on forecasting of maritime trade in 2019, the logistics industry was already moving full-speed to collect data on events in business operations. Yet, the inability to apply Machine Learning on the data streams entering the companies’ data warehouses was slowing down digitization in the industry.
Zuzanna quickly turned to her friends from the AI and Competitive Programming community saying “Guys, there is 80% of all international trade on those ships, if we help to fix it even a little, the impact will be huge, for everybody”.
Jan was just back to Europe from Google Brain and jumped right on board. So did Adrian, Claire, and Bartek. We started to work with the leaders of the logistics industry. We put our heads down and learned the business, the logistics operations, the complexities, the enterprise processes and software architectures…. Guided by our incredible clients.
We quickly realized it was not an issue of missing Machine Learning models, nor just a question of putting models into production.
The problem the industry faced was a lack of software infrastructure capable of doing automated reasoning on top of data streams, in real time.
Supply chains necessitate complex data processing logic to determine what constitutes "normal," "an abnormality," and the "new normal" in terms of business operations. This is something that the COVID pandemic and other black swan events like the Suez canal blockage have brutally brought to light.
At the same time, the Machine Learning world was offering models with promising outcomes for one-off, static data snapshots.
We needed to make these models work on top of ever-changing inputs, reflecting new events and updates to the business processes, on a daily or even minute-to-minute basis.
Our answer was to help software engineers to focus on developing code logic, without having to worry about the way data changes over time. We did so by making the data processing framework itself quietly take care of all data updates. This means data products which are easy to develop and maintain, run on top of an efficient engine.
If you use a spreadsheet, you don't want to worry about how it updates all its data columns when you enter a new value in cell A1. As an engineer you don't want to worry about such re-computations when you have a complex data pipeline. What you do care about is for your user to see an up-to-date Business Intelligence dashboard, in real time.
This is the essence of reactive design.
So, we created the first Python framework to easily build realtime and reactive data products at scale. We created Pathway.
Today, Pathway counts 20+ Team Members across Europe and North America. If you would like to be part of our adventure, just let us know! Join us!