1Mby1M Virtual Accelerator AI Investor Forum: Ihar Mahaniok, Geek Ventures (Part 4)

Sramana Mitra: Let me ask you a broader trend question based on this case study you just took us through, which is really exciting and interesting. There is a lot of talk right now in the industry about full vertical AI solutions in various industry spaces. How do you see that playing out?

There are also examples like General Catalyst doing private equity deals with vertical solution companies. Their thesis is to transform these companies by integrating AI throughout, and then take it from there — that’s their vertical AI thesis.

At the same time, there are vertical AI products being developed and sold to those kinds of companies. There are many variations of the vertical AI thesis floating around. What is your vertical AI thesis?

Ihar Mahaniok: I would separate this into two buckets: vertical AI software and AI services.

The main difference is that vertical AI software is sold to businesses that use it to deliver their services. For example, if a company is developing AI voice agents for healthcare, they aren’t providing healthcare themselves — they sell voice agents to healthcare companies. That’s vertical AI software.

It’s an evolution of software-as-a-service (SaaS), but with a much stronger focus on efficiency and cost savings. This is a big part of what we do, and it’s important that these companies understand their clients. Traditional SaaS approaches like payment by seat or usage still apply here.

The second bucket is AI services. In this case, the client buys a service, not software. The key point is that clients don’t really care whether the service is performed by AI or humans — they just care about the results.

A clear example is AI recruiting services. A recruiting agency might use humans or AI to find candidates, but as a client, you only care that you get qualified candidates. That’s different from vertical AI software.

Sramana Mitra: It’s business process outsourcing — AI business process outsourcing.

Ihar Mahaniok: Exactly. Vertical AI software is an evolution of SaaS, which has been venture-backable for a long time. PandaDoc and People.ai, for example, are SaaS companies that have been around for a while and are increasingly becoming AI-first.

Historically, services companies weren’t venture-backable because they required large operational and labor overhead, limiting their growth potential. Now, AI services companies can scale faster since they need fewer people, making them more appealing to investors.

This connects to the story you mentioned about General Catalyst and other VC firms doing roll-ups — they acquire companies and transform them through AI.

Sramana Mitra: They are two different categories, though. The roll-up story and the services story are not necessarily the same — they’re still different buckets.

We’ve covered a lot about the services part. In India, for example, because of the strong IT services background, there’s a natural fit for AI services. Many interesting companies are emerging that provide AI-powered solutions as a service — not just selling software, but delivering complete solutions built on their own intellectual property.

Ihar Mahaniok: Yes, exactly. Let’s not go into roll-ups — that’s not our focus — but AI services definitely are.

If you go back to the company I mentioned earlier – Cytronic. This is also AI in robotics services. Their clients pay for fulfillment, not for the technology itself. The clients don’t care whether a human or a robot fulfills the order — they just care that it’s done efficiently and affordably.

Similarly, I see huge potential in AI services across many industries — accounting, legal, and more. Anywhere services are traditionally delivered by humans can now be automated.

Right now, this mostly applies to white-collar office work, but eventually AI services could move into front-office functions as well.

This segment is part 4 in the series : 1Mby1M Virtual Accelerator AI Investor Forum: Ihar Mahaniok, Geek Ventures
1 2 3 4 5