Q: What makes you lean forward in the first five minutes of a B2B SaaS pitch?
Clarity of problem and founder depth. Early on, I want to feel that this person has lived the problem and that they’ve spent time in the weeds with customers and can articulate not just what’s broken, but why it’s stayed broken. The second thing is depth: does the founder have a point of view on where their sector is going, not just where it is today? In my experience, founders who build things customers actually need do so by combining deep domain expertise with technical depth. In my opinion, that combination is hard to fake, and it is usually apparent early in discussions if founders lack in either area.
Q: Is the current SaaS-pocalypse going to see many SaaS businesses experience their Kodak/Blackberry moment?
I don’t think AI spells the end of SaaS, but it is forcing an evolution, fast. I believe that SaaS businesses most impacted by AI’s evolution are those whose core value proposition is based on automating simple workflows that an agent can now handle directly. But SaaS platforms that serve as systems of record (for example, housing data, business logic, and compliance workflows) have the opportunity to become a foundation for agents to build on, not the thing they replace. Surviving this shift requires more than a product roadmap update; it means fundamentally reconceiving the business including who the user is, how value gets priced and how the product is architected. The companies that can thrive will have genuinely transformed into AI businesses, they won’t be just SaaS businesses with an AI layer on top.
Q: Having co-led PolyAI’s Series D, what did you see in the business and leadership of Nikola Mrkšić?
When we first met Nikola and the PolyAI* team in 2022, we ran close to 100 calls to pressure-test PolyAI’s bots against bots by their competitors. The verdict was clear to us: by our assessment, PolyAI’s bots outperformed their competitors. Equally, Poly had strong commercial substance and had built a platform that global enterprises like FedEx and Marriott trusted for mission-critical interactions. Nikola understood that winning in enterprise voice AI would likely require a full-stack solution: proprietary models, brand-matched voice quality and the forward-deployed engineering depth to make PolyAI work inside complex customer environments. That combination of technical conviction and commercial pragmatism is what compelled us to re-invest in PolyAI multiple times over the past 3.5 years.
Q: What is the self-deception you need to start a company?
Starting a company is a brave action. Most new ventures don’t make it, and even the most successful of companies have extreme highs and lows along the journey. The founders that I’ve been fortunate to work with represent businesses operating in many different domain spaces, but some of the shared attributes I’ve seen include a rare level of optimism and a deep conviction about their specific point of view on the world. In a job where you get told no far more than yes, I think these are essential parts of the founder toolkit.
Q: What technology trends and markets do you believe will be resilient as AI’s footprint expands?
A few areas that I believe are positioned well as AI’s footprint expands include the broad domain of Physical AI which encompasses robotic applications, autonomous systems, and AI applied to real-world scientific and engineering problems. Another area we are excited about is the technology democratization thesis, represented by companies like Replit* that are helping non-technical users become creators of software and that are expanding the total addressable market of the software industry. And finally, I think that companies that support the proliferation of AI, like Render*, which builds developer infrastructure that supports AI-driven code proliferation are likely to continue to see tailwinds from AI.
Across these categories, the businesses with proprietary data, deep workflow integration and strong domain expertise are best positioned to build durable moats in the AI era.
*Georgian portfolio company
