1Mby1M Virtual Accelerator AI Investor Forum: Warren Packard, AI Fund (Part 4)

Sramana Mitra: Interesting. It’s actually highlighted a lot of points that you are willing to cross geographies, which I was going to ask you about anyway.

Let’s do a couple more case studies.

Warren Packard: Another company that I can highlight is a company called Profitmind. This is a company that started out of one of our very first companies we founded. We founded a company called LandingAI in the computer vision space. The idea was to apply machine learning to typical manufacturing situations where you need to do computer vision, you need to look at the quality of products, look at deformations or imperfections, identify those, pull them off an assembly line, repair them, et cetera.

Out of this company, we started leveraging the machine learning to maximize the profit in retail establishments, which sounds completely orthogonal to industrial computer vision. In a way it was, but generalized machine learning skillset can be applied absolutely anywhere.

Sramana Mitra: All places.

Warren Packard: Of course, we looked at this and thought we need to separate these ventures. We’re finding success in industrial applications with Landing AI. We’re finding success in retail establishments with Profitmind. So, we pulled it out. Dr. Mark Crystal was at the helm of Profitmind. He had been in retail for decades and knew the industry well.

In fact, this underscores an important point of how we start companies.

At AI Fund, we are not subject matter experts. We’re not vertical industry experts. We know AI, and we know how to start companies. We don’t know healthcare. We don’t know retail, we don’t know salesforce automation. But by working with a founder in residence, or a CEO who knows these domains, we can very intelligently approach these markets.

This was the case with Mark Crystal. He knew retail very well. We pulled out the company called Profitmind. Basically, it can scour internal information and external information to maximize the profitability of a retail corporation.

This sounds very generic, of course, there’re quite a few companies that want to set out to maximize profitability of retail. But what we found out here with Profitmind is that they were much far advanced than any other technology company. In doing so, we were actually struggling to figure out the appropriate business model for this company because the effects on a business are so extreme and so quick that you can’t possibly price that in. The benefits occur very quickly.

Sramana Mitra: What were they doing that it was so dramatic and so quick.

Warren Packard: This will get beyond my specific knowledge of Profitmind. My role at AI Fund is to oversee all the 46 companies. The one thing that I can say is that the technology that is behind Profitmind predates LLMs or large language models. It predates ChatGPT. It is founded on traditional machine learning and deep learning models. When you’re training a model for a specific endpoint or end goal, in this case profit maximization, there is a bit of uncertainty on how that model is doing. What it’s doing. It is a bit of a black box because you have all those internal nodes inside a machine learning model.

So how is it actually doing it? How was it trained? I would have to tap the AI engineers of that company. But fundamentally, part of the secret sauce is data acquisition. If you’re pulling in a breadth of data, not just internal data, which of course is kept extremely confidential and private for that organization, but also external data that’s available by looking around the web for competitive offerings, pricing, comparing one model to another.

So, you have a complete picture of a market. When you have a complete picture of the market and you understand the pros and cons of the capabilities of a certain organization – profit maximization, revenue maximization, customer retention, those variables can be maximized extremely rapidly in an extremely powerful manner.

So, it’s been a really fun journey to work with this company. As they go into a client company, they do have to integrate with the backend so that they get this data feed. Once they’re able to integrate with them, the results are stunning, which is very rewarding because it shows the power of AI, not just large language models, which are extremely powerful, but also more fundamental machine learning and deep learning, supervised learning models.

Sramana Mitra: This company sounds really interesting. Are they beyond $5 million in revenue?

Warren Packard: They are getting there. The quick answer is no, but they are very quickly approaching that number.

Sramana Mitra: This one sounds really interesting, but I’m sure you have many companies in the portfolio that are equivalently interesting.

As you know, we do these Entrepreneur Journey series of companies that have at least $5 million in revenue. And we do a detailed study of how they’re doing, what they’re doing, and it’s a very big coverage profile, kind of like a feature story type of coverage.

So, if you want to send in introductions and send them in, we’ll be happy to do the stories.

Warren Packard: That sounds great. Thank you.

This segment is part 4 in the series : 1Mby1M Virtual Accelerator AI Investor Forum: Warren Packard, AI Fund
1 2 3 4