This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.
A coffee with… Zuzanna Stamirowska, founder, Pathway
Zuzanna Stamirowska founded Pathway in 2020 after a varied academic career which involved attending a school for French politicians, a military engineering academy and a stint Stockholm studying management consultancy.
Leading experts in the field of artificial intelligence make up the Pathway team, including CTO Jan Chorowski, co-authors of Geoff Hinton and Yoshua Bengio and business Angel Lukasz Kaiser, who co-authored Tensor Flow and is known as the “T” in ChatGPT.
TI caught up with Stamirowska at this year’s Big Data London event, where Pathway announced a new Python-based app for creating real-time AI applications that continuously learn from proprietary data sources.
There are a lot of data tools out there, how do Pathway’s help enterprises?
Mostly they reduce waiting times for results and enable intelligence that wasn’t possible before — for instance, analytics involving complex batches of data. So, you can get real-time data to get results and insights immediately instead of dealing with data that is old.
What are some real-life applications?
We have a strong footprint in logistics, where there’s masses of data and global supply chains, and you need to understand where your packages are. La Poste uses Pathway to optimise its operations and we’ve managed to reduce its CAPEX spend by 16% in data management, by being able to process data intelligently, with low latency.
The logistics involved the transportation of maritime shipping containers to orchestrate workflows is another use case. Over 80% of all international trade travels on ships, When the Suez Canal was blocked for six days in 2021, the questions companies were asking were ‘What are the affects of this blockage? Where should I reroute my ships?’ These are places where our technology can really make a difference.
And in financial services it is also used to detect fraud and analyse transactions.
So, your tools are good for real time or near real time situations…
Yes, but not all use cases are about very short latency where you need the results immediately. There are other applications that take so long to compute that people do them in weekly cycles and then must wait for hours for calculations. It means people can get results in a few minutes rather than a few hours.
How do LLMs and ChatGPT interface with Pathway?
We can help engineers and developers using LLMs for real-time decision-making of proprietary data: They can be used to structure data, for information retrieval out of masses of enterprise data, they can speed up things enormously.
One of the concerns with enterprise use of LLMs is that control tends to go outside the enterprise and there are doubts around privacy and liability of the systems.
Enterprises want to keep their databases closed and they want to minimise the number of external bits. And this is where we come in, because with Pathway you can use an LLM model, wherever it is, and keep everything within your own ecosystem Our tools are the marriage between real time and LLM which offers a very powerful application layer that allows you to do more complex things.
What was your pathway to Pathway?
I went to the Panthéon-Sorbonne French school for politicians and then L’Ecole Polytechnique, a military engineering school. I also spent some time in Stockholm studying management consulting and game theory. It was there that I fell in love with game theory – so much so that I met John Nash – who the movie A Beautiful Mind was based on – and I went to one of his lectures.
How did that shape your career?
Game theory for me was the real switch between politics and a more scientific path. I specialised in game theory on graphs, which has applications mostly in transportation – modelling and predicting what is going to happen in transportation systems. And I did my PHD on forecasting on maritime trade which was published by the National Academy of Sciences in the US. This was how I got to understand about data and its complexity.
When you’re not studying data how do you recharge your batteries during the working day?
I have strategies that I have developed over the years to make sure that I recharge quickly. I have a beautiful veranda that is filled with exotic plants. And very often I don’t sit at the desk all the time I go and have a short walk.
I think it’s sad that people have lost the habit of purposeless walking, some of the best ideas I’ve had have come to me when I’ve been walking beside someone and just talking. This is the best way to get innovation out there.
#BeInformed
Subscribe to our Editor's weekly newsletter