This is the fourth in a series of interviews I am conducting with members of the Machine Commons supplier Collective. Subscribe to the site to be alerted about future posts, or become a partner today!
Przemek ‘Shemmy’ Majewski is the CEO and co-founder of DLabs.AI, an AI solution agency based out in Poland.
How’s life in the pandemic era? Have you or the business felt much impact?
“The last few months were particularly harsh. I’m trying to keep focused on moving forwards.”
What do you do?
“We’re an agency of machine learning experts working with customers across many different sectors and applications – our focus in in deep learning, natural language processing (N.L.P.) and reinforcement learning. Computer Vision (C.V.) is our strongest capability at the moment.”
“We have a network of experts, some part of the company and some supporting as vendors.”
By using contractors, how do you find the project management ?
“Ah, very good question. I don’t personally manage it, but it works well at the moment.”
No missed deadlines? Do you ever have issues with production not delivered on time?
"Yes. We're in the space of machine learning where not everything is predictable(!). This is why we focus on small wins as fast as possible, so we can show value and transparency to build trust."
"When there is trust, there is also more freedom and time to work and deliver outstanding work."
Why is C.V. the focus?
“Customers started asking about that, so we thought we should invest in it! And I think it's a good step because Computer Vision can bring many obvious benefits to companies, such as simpler and faster processes, and the reduction of costs.
"Moreover, computers are very accurate, and, contrary to a human eye, they never get tired."
"Computer Vision has a wide range of applications from medical (like eye disease recognition) to industry (like autonomous driving) or even marketing, so I think it will be gaining popularity in the next few years.”
Fair enough. Success stories in mind?
“Many very different projects… one example is a monetized gaming company, our implementation helped them raise a lot of funds.”
His win is when the client wins.
What’s your role in all this?
“Whilst I have a technical background, I work in business development – I love finding solutions."
"It’s so great to work at the intersection of and connecting people and technologies.”
How do you build your client base?
“Mostly started from networking – these were the main sources of deals for a long time. At this stage we produce a lot of content, which is gaining traction and customers are now coming to us. Finally starting to pay off this effort.”
What type of content?
“Our last e-book was how to build C.V. system step by step.”
Seems a little technical for clients…
“Well, yes, but not only this; we also write a lot of business articles, about how AI can help.”
From Towards Data Science, ‘Ten reasons your company should implement AI.’
What got you interested in machine learning?
“I had diabetes for over 25 years. It’s the kind of condition that affects life in many different ways. You have to count how much you eat, how much insulin you take, your glucose level, physical activity... all these factors are interconnected.”
“I was working at the CERN Swiss nuclear facility and it was the first time I encountered big data and machine learning tech."
"I figured out that having a lot of data with this kind of tech might help me build a solution for people with diabetes to better manage their condition.”
“I won some acceleration programs and got funding, but discovered that the problem was so big that we needed too much funding. We thought ‘why not also focus on other projects’.”
What’s your approach to solving problems?
“The most important thing when talking to a customer is to really understand business objectives. Why the ML solution? How will it make the business better? Saving time, resources? To generate revenue?”
“Every milestone and KPI needs to be geared towards the business metrics.”
What are mistakes an engineer might make from a BD perspective?
“The classic mistake is not checking if part of the solution is already somewhere on the market.”
“A customer might come to us and say ok build me a C.V. system that detects faces… but to save engineering time and client investment… we already know of at least two solutions that exist on the market!”
How does the client charging work? What about the license?
“Some customers need or want to hold full IP rights for solutions (banks are a good example)."
"For many others, we can usually use i.e. Google APIs and generate the solution faster than building it from scratch.”
What are you excited about for the future?
“When we started five years ago, it was very hard to explain what this technology could actually do for the client."
"Now, people are coming to us with existing ideas about what they want to do.”
“I’m excited about how technology is digitizing the world. Every day we read about another company solving another problem. Where is the limit??!”
“I love seeing the kind of startups that keep popping up.”