Employee Spotlight-Alex Ivanichev Sr. Director, Software Engineering, Brand Safety & Classifications

Where do ideas come from – particularly when it comes to our engineers at DoubleVerify? We had a chat with DV’s Alex Ivy Sr. Director, Software Engineering, Brand Safety & Classifications, to see how he is able to come up with new and strategic ideas to help bolster DV’s technological capabilities.

1. How long have you been working at DV?

I have been working at DV for 2 years.

2. Can you tell us about your position and about your Group?

As Senior R&D Director, I’m leading the Brand Safety & Classification group. My group consists of both product and platform teams, and we are responsible for over three major fields:

• Real-time traffic system that processes 100B+ events per day with up to 15ms latency
• Classification - we need to understand and classify multiple types of content (text, images, video, and audio)?
• Platform - teams that are responsible for building infrastructure to improve the delivery and increase velocity

3. Can you walk us through some of the growth that led you to your current position at DV?

I started my career as a full-stack engineer and was building systems end-to-end, building both frontend and backend – configuring servers, deploying the code, everything that was required to build and deliver the system from the requirements stage to production. I was fascinated by taking multiple disciplines and various building blocks and consolidating them into fully working solutions. During this time, I was working with many other developers, helping them with their tasks and approach to problem solving correctly. I understood how much impact I had, which was way beyond the scope of the single project and it helped me understand that being a manager was my true passion.?

From there, I began leading people and helping them grow both technically and professionally, and that helped me achieve a multiplier effect way beyond what I could do as an individual contributor. From there, I held many different positions: team lead, senior team lead, group leader and VP R&D.?

4. What attracted you to work at DV?

Over the last few years, I have been focusing on how to achieve more from a combination of Big Data with AI. That is why I was so passionate about joining DV. At DV, I have the opportunity to lead teams that are responsible for multiple disciplines, and we have a huge amount of data, processing more than 100 billion events per day and using ML/AI to make smarter classifications.

I joined to make our system much more intelligent and scalable – that way we can deliver more value to our customers at a fraction of the cost.

5. Can you tell us about a project that you learn from?

One of the least successful features was related to our backend infrastructure. Our engineering teams continue to deliver dozens of new product features, which requires them to build many new components: micro-services, streaming pipelines and various ETL jobs. We started to see that in every new project we invested at least 50% on building and preparing infrastructure.?

Additionally, we noticed that multiple teams may have invested in similar solutions, which led to wasting time. Our immediate reaction was to try to make as much reuse as possible and shareinfrastructure/solutions that the team had created with other engineering teams and, basically, we gave the teams the responsabilty of owning infrastructure along with the product work.

That solution proved to be very effective and we were able to reduce 20-30% of the effort.

However, as multiple teams adopted the infrastructure of other teams, multiple problems started to arise.

To solve the challenges, we decided to create a new team –? the ”backend platform team” – that provides infrastructure as a service for other engineering teams and provides them both generic enough solutions for their needs.

The team provides a level of support that will allow them to bring new product features to production fast.?

6. Can you tell us about a very successful project at DV that you feel proud of?

I will describe one of the projects that we pushed that really had an impact on DV business. One of the areas that I’m responsible for is our classification system. As part of the classification flow, we are integrating multiple social platforms like Facebook, Twitter and TikTok into our system– originally because every platform is different in the content it offers to the end user.?

We had multiple customizations for every pipeline, and this approach worked well for us for some time. But, as our business started to grow, we began to onboard many new platforms to the system. The cost of supporting new platforms required a lot of effort from our engineering side, and, as you know, necessity is the mother of inventions. Hence it led us to completely rethink our system and use what is called a “platform approach.”?

We designed our system from the beginning – when we have built a level of abstraction over the content (text, images, videos) and separated the classification engines from the content by making it classify content and return low level terms that then can be aggregated together based on the platform policy, along with other benefits like human input, caching, etc. This approach allowed us to achieve separation between the platforms, reduce cost (because of the unified caching) and, most importantly, support and onboard new platforms in hours instead of weeks of work while still preserving the customization of each platform based on policy.?

7. What made you want to be in a leadership role?

When I first became a developer, I was mostly passionate about coding and learning as many new technologies as possible, as well as how to combine them together to build a product.?After some time, I began to develop expertise in those technologies and programming languages. At this point, people started consulting with me on “how it is better to implement something,” they all had worked on various projects, supporting different use-cases, it was very interesting and challenging to help them with their work, I think this was the first time when I realized that no matter how well I develop and how fast I can write code, it is still limited to scope of one project compared to the impact that I can make helping others to grow and pointing them in the right direction I can achieve the multiplier effect.

8. In your opinion, what qualities does a great leader need?

There are a lot of qualities that a great leader must have such as: integrity, great communication skills, vision and a results oriented mindset. From my personal experience, great leaders must empower the people they lead – because even if you’re a great developer but completely missing the end goal of your task you will never be able to provide a good solution. If you won't be a part of the discussions, you will not have the information and you will be unable to solve the problems fully. If you always depend on others to tell you what to do, then you won’t be able to make a decision on your own. It is really important to help the people that you lead by giving them the necessary knowledge to make their own decisions, which I believe is essential for great leaders.

9. What tip would you give to people who would like to work as developers or team leaders??

You must have a good understanding of the system and the product that you are working on. I saw a lot of examples where developers and team leaders that were focused mostly on technical skills miss the reason for developing. They think about the feature or how the customers will use the feature that was required to develop. This often leads to over-engineered solutions or premature optimization, instead of trying to understand the reasoning and provide a minimum viable solution. Collect the feedback and then reiterate and improve it.

10. Why do you think the field of big data is in such high demand today?

Today, data is everywhere – data is the new gold. Data is being generated by everything: from watching a new movie on Netflix to listening to a new song on Spotify. To process this data and take insights from it is very challenging. This is where AI comes into play. We're able to learn our usage patterns and provide recommendations, detect anomalies, and more. So, I definitely expect the data field will continue to evolve.

11. Which technology is a must for any software engineer to understand?

There are many different technologies out there – Cloud, Kubernetes, ML – but if I could suggest one, I would pick Kafka. I think Kafka is one of the foundations for building driven architecture and a highly scalable system. Kafka leads to building more scalable and resilient systems.