At a time when conversations around AI are evolving almost daily, the challenge for businesses is no longer deciding whether AI matters, but understanding how to approach it responsibly while the landscape is still taking shape.
Every week brings new tools, frameworks, and predictions about what may transform the advertising industry next. Between the excitement, the hype, and the genuine breakthroughs, it can be difficult to distinguish what creates immediate value from what still needs to prove itself in practice.
At Virtual Minds, we believe the most sustainable path forward is neither blind adoption nor passive observation, but disciplined exploration grounded in real-world value. In this article, Bikram Shrestha, Lead Product Manager – Data Science & AI at Virtual Minds, shares how we approach AI at Virtual Minds: building on years of embedded intelligence in our platforms while carefully evaluating the next wave of technologies shaping the market.
In this article, Bikram Shrestha, Lead Product Manager – Data Science & AI at Virtual Minds, shares how we approach AI at Virtual Minds: building on years of embedded intelligence in our platforms while carefully evaluating the next wave of technologies shaping the market.
From embedded intelligence to emerging AI
Intelligence has been part of our platform DNA since the beginning. As the technology landscape evolves, our philosophy remains consistent: deliver measurable value today while preparing for transformations that may reshape our industry tomorrow.
At Virtual Minds, AI isn’t a separate initiative. It’s woven into how we build products and serve clients. The algorithms and data models powering our product platforms have been continuously evolving since 2001, quietly making advertising technology smarter and more effective.
Long before Generative AI became a boardroom topic, intelligence was already embedded in the way our systems optimized workflows, improved decision-making, and reduced operational complexity. Now, as Generative AI and Agentic AI capture industry attention, we’re approaching these developments with the same mindset that’s guided us from day one: clear-eyed optimism paired with pragmatic execution.
We believe innovation creates the most value when curiosity is balanced with discipline and experimentation is grounded in practical outcomes.
For our clients and partners navigating the same landscape, we hope sharing our thinking is useful; not as a blueprint, but as one perspective from a team that’s been building in this space for a long time.
Intelligence built in
Our platforms have always used data, algorithms, and automation to solve complex challenges. We didn’t always label these capabilities “AI”. Many emerged as machine learning, predictive analytics, or optimization algorithms, but the intelligence was real, and more importantly, it was useful.
What matters to us is not the label, but the impact that these technologies have on real-world operations. These are not experimental features but more production systems our clients use daily to make better decisions, optimize performance, and reduce manual effort.
The value lies in intelligence that is deeply embedded into workflows, not layered on as a feature after the fact. That foundation gives us a perspective on what is genuinely transformative versus what is merely novel. It also gives us a healthy level of skepticism toward technologies that look impressive in demos but remain unproven at scale.
The hardest part of navigating AI right now is not the technology, it is knowing what to resist.
Between potential and practical testing
New standards and frameworks including Model Context Protocol (MCP), Ad Context Protocol (AdCP),Agentic RTB Framework (ARTF) and others, are emerging that could meaningfully change how ad platforms integrate and communicate.
We are not watching from the sidelines. We are actively exploring real-world use cases to better understand what these protocols can actually deliver, as opposed to what they promise.
The opportunity is significant: lower integration complexity, more adaptive systems, and entirely new forms of automation across the advertising ecosystem. Our view is both optimistic and pragmatic. We see genuine potential in technologies that could reduce integration complexity and enable new forms of automation. At the same time, we recognize that not every promising technology will deliver lasting value, and not every emerging standard will achieve meaningful adoption.
History shows that early momentum alone is not enough to guarantee long-term relevance. This is why we emphasize controlled exploration over early commitment: learning systematically through contained use cases, making decisions based on evidence rather than excitement, and preserving the flexibility to move fast when the right opportunities become clear.
While our commitment to practical AI value remains unchanged, the possibilities around platform interoperability and automation are evolving rapidly. Our teams are actively exploring AdCP, MCP, and related protocols to understand how they could reshape integrations, scalability, and operational workflows across the advertising ecosystem.
The momentum behind these developments is real. That is precisely why now is the right time to be learning, while the standards are still forming.
For Bikram Shrestha the challenge is not simply understanding what AI can do, but recognising when restraint is equally important.:
“The hardest part of navigating AI right now is not the technology, it is knowing what to resist. We don’t claim to have all the answers on where AI takes this industry. But the cost of getting it wrong is real, and that is why we stay close, experimenting and asking the right questions, carefully, and with our clients’ interests in mind.”
That balance between curiosity and discipline informs how we evaluate emerging technologies and guides the principles outlined below.
What guides our approach
First, continuity matters.
Intelligence has been central to our product architecture since the beginning. We’re building on a foundation of experience, not starting over with each new wave of innovation.
Second, we build for a measurable impact.
We invest where we see clear paths to creating measurable value, reducing costs, improving performance, or enabling capabilities that weren’t previously feasible. Equally important, we are willing to stop investing when the evidence no longer supports continued effort.
Third, we stay grounded.
We explore new technologies through controlled, disciplined experiments. We celebrate genuine breakthroughs while remaining focused on what creates real value for our clients.
The path ahead
The advertising technology landscape will keep moving. Some of what’s emerging today will reshape the industry. Some will not.
Our responsibility is not to predict the future with certainty. It is to stay close enough to the developments that matter, learn quickly enough to act when clarity emerges, and protect our clients from the cost of moving too early in the wrong direction.
That balance – between curiosity and discipline, innovation and responsibility – will continue to shape how we approach AI moving forward. We will continue sharing what we learn. Not because we claim to have all the answers, but because these are questions the industry will need to explore together.
