1Mby1M Virtual Accelerator AI Investor Forum: With Jukka Alanen, Rebellion Ventures (Part 5)

Sramana Mitra: We’ve done vertical industry-specific problems and functional problems in this discussion so far. Could you discuss a platform company that you have invested in?

Jukka Alanen: Yeah, there are actually quite a few. It’s interesting that many of these companies may not necessarily start fully as platform companies; they start with one beachhead vertical or a beachhead use case. Then they expand and broaden more into a platform company over time.

So I’ll give you two types of platform companies.

So one company is called Ubicept, which is a spin-out of the MIT Media Lab. They have a new type of computer vision technology that can radically provide better quality. Especially in today’s computer vision issues, if you have fast movement or poor lighting conditions, the computer vision doesn’t really work well. It affects computer vision in real-world use cases, especially in verticals like logistics, automotive, or transportation. They have a brand new technology that they pioneered that radically looks at things at a single photon level, and they can create very different quality of computer vision. That’s an enabling technology that can then be used in a number of different verticals.

Sramana Mitra: So in the images that computer vision captures, it augments those images by applying machine learning.

Jukka Alanen: Yes, they are essentially analyzing light at the level of single photons, and then they use AI and ML to process individual photons. Then based on that, you can have as good imaging perception and vision as nature allows, because you’re operating at a single photon level that you’re then analyzing. That’s a perfect use case for AI because you have a very large number of photons for which you need AI and ML to be able to analyze and make sense of all of that data.

Another example of a very different platform company called HuLoop Automation. They provide a platform for process automation for small and medium sized enterprises. They’ve focused recently on financial services and retail, but it works across multiple industries. The issue with process automation that many of these companies face is that many of the traditional process automation platforms were defined for large enterprises. So they’re very complex, very time consuming, require a lot of provisional services or expertise, and take a long time. They’re not necessarily a good fit for small and medium sized companies. So, they built a new type of platform that enables multiple types of use cases and applications in multiple verticals, whether you’re a bank or you’re a retailer to automate a number of business processes from loan collections to how you manage your catalog for retail SKUs.

What I’m seeing in the practice is that there are many different types of platforms, and they’re platforms on top of platforms. A hierarchy or stack of platforms is emerging in this industry. Many of those platforms use other platforms. It’s like a collection of different platforms at different abstraction levels, depending on what is needed to add value for customers.

Sramana Mitra: What I see is in all of this, not in the computer vision example that you gave, that is a very, very specific technology problem, but the rest are all domain specific knowledge that is being codified with AI that can automate functions and then you build layers of abstractions on top of that. So whether it’s sales intelligence or retail catalog building or insurance agents, it’s basically really understanding the step-by-step process of the domain enough to be able to apply AI and automation.

Jukka Alanen: I think that’s a really important point because it’s very hard as a brand new startup to say that we’re gonna be a very broad platform that does all things to all people. That doesn’t really work. When you’re a startup, you have to be focused and have your beachhead use cases. You may already have from the get-go a platform vision in terms of how you’re architecting the platform, how you think about data, and how you think about your ecosystem. It’s good to have a vision, but you need to be, on the other hand, very pragmatic in terms of where can you solve customer problems and start getting adoption and building out market traction, as well as get revenue. You need to be very focused and have a beachhead approach there.

This segment is part 5 in the series : 1Mby1M Virtual Accelerator AI Investor Forum: With Jukka Alanen, Rebellion Ventures
1 2 3 4 5 6