In this interview, we have the opportunity to hear from Oleg Lola, founder of MobiDev, a software engineering company that specializes in helping midsize enterprises integrate AI into their products. Oleg shares his background and journey, highlighting his passion for technology and his goal of serving visionary entrepreneurs by adding value to their product through technical solutions. He also addresses the misconception that AI is only accessible to big companies emphasizing MobiDev's focus on making AI technology easy to use for midsize enterprises in the U.S. Let's dive in!
U.Today: Oleg, please tell us about your background. When and why did you start MobiDev?
Oleg Lola: I was born in Mariupol, Ukraine. It was a large industrial city. Bad ecology, tough and smart people. Both of my parents were teachers - with low incomes, they worked second jobs to make ends meet. They stressed good education and striving to achieve a meaningful life. They never forced me, only helped me, with their example and actions. As a child, I helped my parents grow vegetables and sell them in a local farmers market.
Like my parents, I was hardworking and I "swallowed" books - fiction first, but then technical literature. Early in my life I was lucky to get access to computers, which in the early 90s was unusual in Ukraine. I used to read books as a teen on Assembler, C++, and 3D programming and had fun with it. I won several high school national-level competitions in math, science and programming, and created several desktop apps to help students visualize stereometry concepts. I felt like a trailblazer in school. I graduated with honors from high school (Mariupol Gymnasium, Ukraine) and earned a B.Eng. and M.Eng from Kharkiv University of Radioelectronics in Ukraine.
Before starting my company, I considered myself a classic nerd/introvert and was proud of it. Coming from a math and tech background, I never thought I would be talking to people that much at work (laughing). Actually, the best perk that has helped me as an executive is learning how to see an opportunity. In my entrepreneurial career, I have enjoyed change, challenge, and trying new things. Such creativity, combined with nerdy persistence, helped me create my software engineering company in the recession year of 2009, driven by my vision of mobile devices. My vision and creativity helped me establish my company's brand, an ultimate promise of delivery and service, and grow to 400 people by the beginning of 2023. They also helped me to pivot to machine learning and AI services in 2018 before AI became mainstream.
Before starting MobiDev, I did scientific research at my university and worked at other software companies. While I really adored research, which is all about experimenting, finding new options, trying things out and, as a result, bringing new value, I really hated my experience with other tech companies as a software engineer. At that time, outstaffing was really popular, which meant the tech company would just find "brains," connecting clients and engineers, with the hope that clients and engineers would sort things out. It meant that clients had to take on a lot of the management burden by themselves, and the initially exciting development process became too cumbersome. I thought that I could do better as a business owner. Thus I kicked off my software engineering company MobiDev with the idea of serving visionary entrepreneurs and adding extra value to their product. From day one, MobiDev started client interactions with understanding the business problem, researching and offering new ways of finding technical solutions, taking accountability for delivering results and handling thousands of small but important things within the course of a software development project, both technical and human-related.
This way, visionaries could focus on what matters most: their business. Since 2018, MobiDev's focus has been helping build AI apps for visionaries, along with making AI technology easy to integrate into their products.
U.Today: AI has come a long way from being a mysterious rocket science like technology to overhyped public applications like ChatGPT and FaceApp. AI looks like a game for titans. Is it possible for midsize enterprises in the U.S. to utilize AI?
Oleg Lola: Absolutely: AI is heavily overhyped. The launch of ChatGPT made things even worse. There is a misconception that AI can do anything, and robots will soon replace humans. As a professional who has worked with machine learning years before ChatGPT, and as an AI enthusiast, I would say: take it easy. I see AI as another set of technologies for automation and making our lives better. Businesses can benefit from AI in two ways:
- Use products that already contain AI technology. Hint: Search the web for AI products in your domain area, and test drive them on your own according to your specific needs.
- Generate extra value for your high tech product with AI. Hint: Have AI/ML professionals analyze how AI can be beneficial for your product and your users. In the vast majority of cases, you will not need to create yet another GPT. It is better to explore the options of fine-tuning existing models with your data, using pre-trains or even connecting to third-party APIs. The power of AI is about using your creativity and some professional help to make the magic happen.
What is really important is that both options are valid for businesses of any size.
U.Today: How can AI assist companies in resolving their challenges, and what specific problems can be addressed through AI?
Oleg Lola: There are two ways of solving problems in computer science.
- Algorithmic: Where a human defines how a specific problem must be solved, based on expert opinion, science, or business rules. The outcome is exact, strictly defined by the heuristics.
- Machine learning: Where a model learns how to respond to a certain input, based on data. The result here is an approximation. In most cases, it is a very complex nonlinear approximation. So, one might say machine learning models make predictions about new data based on the data it already digested.
Those problems that might be solved with prediction can be targeted with AI, mostly connected with perception and patterns extraction. Taking into account that the cost of prediction (hardware) is becoming lower, there is an amazing opportunity to disrupt, tackle new problems, and find new business models. I can name some problems off the top of my head that we have solved for our client-visionaries, while enriching their products with AI:
- An organizational chart recognition for attorneys, both printed and hand drawn
- Truck size and weight estimation based on stationary AR point cloud (Lidar) data
- Inventory tracking and ordering based on sales forecasts for nightclubs and bars
- Bespoke clothes ordering based on visual inspection and human pose estimation
- Biometric identification for insurtech and healthtech
- Smart contract analysis and classification
- Invoice recognition for healthtech
- Training athletes by recognizing their movements and comparing them to professional examples for the sports industry
- Garbage sorting for recycling
- And, of course, customized chatbots based on GPT technology
I would like to address those organizations that have launched high tech products or are planning to. Maybe you want to work on your internal enterprise product for automation, and you have or have not been collecting data. Maybe you need help with software as a service that you use to provide service to your customers and you want to be ahead of the competition. Maybe you need help with a custom hardware solution where optimization is critical.
In my experience, in 95% of cases, AI brings extra value to the product right away based on implementing the frontier machine learning models. It could be increasing customer satisfaction or improving customer experience, and even providing new features, solving those problems that you always wanted to target but never could.
"Right away" here means there is no need to conduct 1-2 year long scientific research to create a machine learning model.
It has already been done by a team of scientists. At the same time, please allow a meaningful time frame of 1-2 months for building the first POC (proof of concept) of your specific AI solution. This would include an AI consulting stage when the vision of your product is defined, a development stage, and an implementation stage. While the development and implementation stages are something you would encounter in any software development project, the AI consulting stage is different and is probably the most valuable for any business exploring opportunities to embed AI in their product.
U.Today: Can you tell us more about this AI Product Consulting stage?
Oleg Lola: Tech visionaries think big, wish to change the world around them, change thinking and perception, find the drive in customer value and get emotional uplift while generating new value for customers.
Tech visionaries already work on the edge between the problem and the solution. AI Product Consulting helps to broaden the edge.
In a practical sense, it helps an organization see the field of opportunity that lies at the intersection of the current organization's business and product, market demands and competition as well as the ML/AI technologies available out there.
For completely new Al Product Development, we make sure the idea is feasible from a tech point of view and meets real market needs. Our Al consultants prepare businesses for the successful launch of their project by defining the final scope of the solution for the MVP and the go-to-market strategy. Next, our engineers turn this strategy into a functional Al product taking into account all technical and business requirements. For embedding Al integration into existing businesses, we offer ways to outpace competitors by implementing innovations that drive real business value. Our experts conduct a detailed analysis of business goals, datasets, and IT infrastructure to offer the best solutions that fit the existing ecosystem. With 13+ years of experience in software development, MobiDev's tech experts ensure that AI functionality is fully compatible with existing software. MobiDev consultants, who are experts in market research, business analysis and AI engineering, go through several steps that deepen the product's vision: idea brainstorming, product and market analysis, and AI model research and estimation. As a result, the product development roadmap and the budgeted options for building and marketing an AI proof-of-concept are created.
U.Today: Do both MobiDev as a company and yourself, Oleg Lola, utilize AI to tackle everyday issues?
Oleg Lola: We use a broad range of AI-powered products that help speed up the development and deployment process, including Github Copilot, TestGrid, and Cast.AI. We want every dollar clients put in development to be as efficient as possible.
U.Today: What are your future plans and goals?
Oleg Lola: At MobiDev, we will continue our laser-sharp focusing on serving visionaries, making their journey of software engineering light and enjoyable. We believe that within a few years, more and more organizations will turn to AI technology and the capabilities of AI will only grow. So, we see a combination of AI and the very best service standards of MobiDev professionals. We will add extra value to visionaries and their high-tech products. We work the best with established software companies that want to bring more value to their existing product, launch new versions of their existing product or launch new products. We often recommend embedding AI and offer options to our clients. The visionaries, in turn, see the options and make decisions that suit their vision best. We are constantly working toward this future, so we clearly envision it. We will keep building game-changing products. We will continue to adhere to MobiDev's philosophy and take care of technology and project management down to the smallest detail, allowing our customers to focus on business strategy and product vision. Our pipeline of generating and implementing breakthrough ideas will continue to run smoothly.
To do this, we are currently investing 30% of MobiDev's engineering efforts into new technology research.
U.Today: What aspect of MobiDev are you most proud of?
Oleg Lola: First of all, I am really proud of our clients. Working with visionaries for over a decade, I see that true vision and hard work bring awesome results. I see our clients going to the next levels of their businesses, growing nationally and internationally, making strategic alliances with major industry players and enjoying the results. I am grateful to all of our clients fortrusting MobiDev to handle their tech needs for many years. I am also proud of my team at MobiDev, the majority working with us for years and bringing their passion and hard work so valuable to our clients. I am proud of our mutual support and look forward to their professional and career growth in the years to come. Within the course of these years, we have also done a lot of good for society. We have launched tech educational courses for single moms, youth and veterans, have organized international conferences for professionals in software engineering, have launched several free apps with millions of downloads (the latest being an AI app to help sort the garbage for recycling), have actively helped the software community to develop with a few open source projects (the brightest of which being one of the first frameworks for cross-platform mobile apps), and have participated in ongoing humanitarian initiatives in Ukraine. I am fascinated by how our work for clients and our active position changes the world for the better.
U.Today: What is your perspective on the future of AI? Do you believe it has the potential to become a frightening dystopia?
Oleg Lola: Well, I see much more of a threat from climate change, terrorists, and nuclear weapons. AI is just a tool. It is powerful, but there are much more powerful tools out there in the world. I definitely support AI regulation. AI must be regulated in terms of how AI products may influence humans and humanity. AI products must conform to privacy protection, copyright, human rights legislation, and other laws. As for the future of AI, we have seen the exponential growth of AI capabilities and we are going to witness more to come. Probably the biggest question one should ask themselves is if they want to be creators of useful AI products, or be a user of someone else's products.
U.Today: Do you have a cryptocurrency portfolio? If so, could you mention at least three holdings?
Oleg Lola: I am rather a homebrew ETF investor. Also, I put some funds into early stage AI start-ups through diversified funds. Blockchain as a technology is more interesting to me than any specific crypto asset. I have a feeling that the decentralization of operations and data-based trust characteristic of blockchain are in line with today's social trends. Products that combine AI and blockchain, for example, are very promising.
U.Today: What advice would you offer to individuals interested in establishing their own business in the AI field?
Oleg Lola: Explore the market, figure out your strengths, see what you can do with AI technology. Go lean, launch early, gather feedback, and reiterate frequently. Find balance between your idea and market needs. Seek mentoring and consultancy, both in business and tech fields (and especially AI).
Arman Shirinyan
Yuri Molchan
Godfrey Benjamin
