Citizen of Silopolis: Dmitry Kan

Dmitry Kan is a Lead AI Scientist, working in particular with search-related projects (read his own Medium for technical posts on the topic). Dmitry – or Dima to his friends – is a versatile person who can be described as a technical leader, founder, software engineer, researcher and entrepreneur. At Silo AI, Dima has been working with one of our large clients, leading the team of both AI experts and software engineers. As a beloved colleague, Dima has been also doing what he always does: fostering a culture of support, trust, and collaboration to build solid teams that deliver quality research and software in time.

Prior to joining Silo AI, Dima spent more than 10 years building the financial search engine for AlphaSense from scratch. Dima was responsible for building a search team and search technology powering the AlphaSense product, which is today used by thousands of reputed banks, hedge funds, and companies in almost any industry vertical worldwide. Most recently, he focused on launching an AI Research Team within AlphaSense.

Dima has also a solid academic background and has worked as a lecturer at the University of Helsinki since 2017. He has a PhD in Applied Mathematics from Saint Petersburg State University, Russia, and a Master’s in Computer Science from the University of Kuopio, Finland.

Dima, in your previous company AlphaSense, you were building a search engine in the financial domain. Now with our clients, you’re building general search engines at a much larger scale. How does this look like in practice?

I’m leading a team of software engineers, owning full stack development areas: frontend, backend and quality assurance. We are tasked with building a scalable search engine involving AI technologies to automatically detect the intent of the user query in multiple languages. The challenge of search is that any query can come in as input, it is not limited to a particular domain. This is a new, fascinating challenge for me.

In your professional life, you skillfully combine different departments, like Product and Engineering, Sales and Engineering, horizontal and vertical teams within Engineering. What do you enjoy in your work?

I enjoy working with smart people, who enjoy tackling hard tech problems. In addition I like to take care of the overall organizing the work of the team: from hiring to assigning tasks as well as unblocking to growth. 

In the current project that I’m working on, I’m obviously excited about the opportunity to reach millions of people with our product, and to have the ability to influence the product decisions. In a project like this, I also like the cross-functional nature of the job: talking to multiple teams and stakeholders

As a data scientist, you combine research with industry grade engineering. How would you describe this approach?

I look for practical solutions. Sometimes a research algorithm can be nice on paper, but does not work directly for the client needs. During my career, I’ve had a chance to build my own products and to work with product managers in the same team. This has given me a broad understanding of the field of search engines, natural language processing and machine learning.

Another thing that I value high is communication. Cross-team coordination and establishing the communication processes is another skill and area of focus for me: No one can survive by only having technical or academic knowledge in today’s projects.

What do you do in your free time?

In my free time I like to read (lately books on emotional intelligence and communication excellence), walk in the forest and cycle. I’m also enjoying cooking by following trusted youtube channels.

Now it's time for our final question, one that we always like to ask at Silo AI. what’s your favorite Silo AI value and why?

Build bonds -- this is so important, especially in the current working environment, when we’re all remote. I’m glad Silo AI has things like weekly research club and remote team lunch, where we get to meet also outside of projects and hang out virtually among peers.

Thank you, Dima!

In addition to his own blog, Dima has been sharing his learnings with BERT and Elasticsearch in the online machine learning publication Towards Data Science.