Over the past several years, Peter Ivanov, Engineering Director at Accedia, has played a key role in building the company’s expertise in Artificial Intelligence, Machine Learning, Computer Vision, and beyond. With hands-on experience in the area, Peter has helped shape how our clients approach AI adoption. In this interview, we sit down with him to discuss the evolution of AI, the challenges businesses face in implementation, and what it takes to build a successful AI strategy.

How has the perception of AI evolved in recent years, and what trends are you noticing among businesses exploring AI adoption?

Peter: Over the past six years, starting long before the generative AI boom, I’ve seen a lot of companies familiarizing themselves with key AI areas like Data Science, Machine Learning, Computer Vision, and Natural Language Processing. They were not only playing around out of curiosity but also starting to actively use some of these components in real world projects. Many of these businesses have stayed consistent with their AI efforts, while others just recently have picked up the pace and adopted more structured approaches to their projects.

On the other hand, companies that only started paying attention to AI during the generative AI wave often remain hesitant. While executives may show interest and the intention to invest, challenges arise at the execution level. Some companies make progress, but others struggle - either due to a lack of proper ideas or difficulty securing budgets. As a result, their AI efforts often stall.

Interestingly, newly starting businesses tend to be more proactive, investing heavily and experimenting actively in the AI field. That’s something I’ve noticed as a trend among our clients.

Overall, the perception of AI has improved, but not as much as one might expect given the optimistic forecasts and general enthusiasm in the business world. It’s a mixed bag: while awareness has grown, the level of adoption still varies significantly across organizations.

What are the key challenges businesses face when turning to Accedia for an AI development partnership?

Peter: I’ve seen a distinct pattern in the types of AI solutions businesses in different industries request from us. In Finance, the majority of requests are centered around fraud detection and customer personalization. Financial institutions face a constant battle with fraud, needing AI tools to quickly detect and respond to suspicious activity while minimizing false positives. At the same time, there’s a strong focus on using AI to personalize customer experiences, from tailored product recommendations to individualized financial advice, as customers increasingly expect services that cater directly to their unique needs.

In media, the emphasis is primarily on content personalization and recommendation systems. Clients want to keep audiences engaged and are looking for ways to offer viewers content that feels curated just for them. The AI-powered recommendation engines that Accedia builds for media clients allow them to suggest the right content to the right user, driving engagement and improving viewer retention.

Meanwhile, in Manufacturing, the most requested AI products in my experience are for predictive maintenance. Manufacturers are keen to leverage AI to predict when equipment might fail, enabling them to carry out repairs proactively and avoid costly downtime. We had a similar case last year when a Swedish manufacturer partnered with us to build a solution forecasting future malfunctions in the production of machine parts. We used Machine Learning and Computer Vision models to recognize and classify damages in uploaded pictures and even managed to scale the solution throughout their factories worldwide.

What do you think are the essential elements of a successful AI strategy?

Peter: Building an AI strategy that truly adds value means recognizing how unique each industry’s needs and challenges are. That’s where we always start off when meeting with new clients. Across all industries, though, one constant is the need for quality data and scalability. Each sector handles different data types, whether it’s financial transactions, IoT sensor data in manufacturing, or behavioral insights in media. Still, one thing remains the same: you can’t build a successful AI strategy on poor data. And while it’s tempting to dive in with advanced algorithms, I’ve learned that starting with a solid data foundation and scalable infrastructure makes all the difference. In the end, effective AI strategies aren’t one-size-fits-all. They’re built with an understanding of each industry’s priorities, whether it’s compliance in finance, operational efficiency in manufacturing, or personalized experiences in mediа.

What would you say to companies that hesitate to invest in AI due to ROI concerns?

Peter: Investing in AI offers significant potential for ROI, though returns can vary based on industry, specific use cases, and an organization’s AI maturity. From what we’ve seen sectors such as finance, healthcare, and manufacturing often achieve high ROI due to their data-rich environments and operational complexities.

Another aspect that should be factored in is the project’s scope. Targeted AI implementations, like process automation, can yield ROI within 6-12 months. In contrast, large-scale AI transformations, such as deploying Machine Learning across entire production lines, may require several years to fully materialize benefits. Still, the results are undeniable - organizations leading in AI adoption see far greater benefits, with top performers achieving a 10.3x return on investment, significantly outpacing those lagging behind in implementation.

However, bear in mind that key factors influencing ROI can also include data quality, infrastructure readiness, and change management. So organizations that start with high-quality data and a clear strategy for integrating AI into existing workflows tend to see higher returns.

How has Accedia developed its AI expertise, and what sets it apart in delivering impactful AI solutions to clients?

Peter: At Accedia, I’ve been deeply involved in growing our AI expertise, and it’s been an exciting journey. What really sets us apart in the market is our ability to make AI solutions practical and accessible for businesses of all sizes. Our AI Capability Center plays a big role in that. It gives us the knowledge and experience to quickly deliver real results, especially in areas like Computer Vision and Natural Language Processing.

Internally, we’ve focused a lot on building our team’s skills. The Innovation Development Center has been a huge part of this - it’s a space where our consultants can experiment and turn their ideas into real solutions. Over the years, we’ve worked on some amazing projects in areas like sustainability, education, and smart cities, which have helped us push our AI capabilities even further. On top of that, we’ve launched internal AI training programs to make sure our engineers stay up to date on the latest developments.

I think that exactly this mix of continuous learning and hands-on experience has really strengthened our position in the market and allowed us to deliver meaningful, high-impact AI solutions to our clients.

Shaping the Future of AI: Conversation with Peter Ivanov

Latest blogposts

Let's Talk