Sramana Mitra: I have another question along those lines. When you are doing AI, typically machine learning and large data sets are involved, what are you seeing in your portfolio or deal flow in how entrepreneurs are accessing that data?
You are talking about taking the technology risks out, but you need data to take the technology risk out in AI companies. What are the best practices that you are observing in your orbit?
Eric Benhamou: Young companies rarely have access to vast proprietary datasets. However, they could make this semi-proprietary by having a unique curation method and having a unique way of optimizing the algorithm to process that data.
Let me give an example. We recently invested in a company originally born in India called AiDash. They focus on processing satellite data imagery. There is so much of it today than five years ago because we keep launching satellites by the dozens per day. There is an enormous amount of high-quality data available that can be analyzed for a specific purpose.
In their case, they used to help companies like PG&E manage their physical assets. By analyzing the data, they can detect whether a shrug or a tree is going to interfere with the power firm or a distribution center. They can optimize it for this use case all the way to the point of automating the workflow of dispatching the product to go clear the overgrowth of shrubbery.
Some of the data is publicly available and some of it is available on license. However, the way you treat this data for this specific purpose is unique to AiDash. The way you incorporate this into the workflow is unique to what we do and the customer group that we cater to.
Sramana Mitra: So you are only focused on the enterprise and not on the small-business-focused companies?
Eric Benhamou: Initially, yes. What happens is that companies who are successful in large enterprises can evolve the offering and go downstream and make the offerings to the SMB. This happens fairly often, but the point of entry and the point of validation is the large enterprise. They are the ones who are confronted with the most difficult problems and this is where you can erect the strongest barriers of entry.
Sramana Mitra: You must have read the articles that I have read recently about the PaaS trend that is picking up. One year ago, it was less prominent, but now it is. I have had conversations with a lot of players who either have lots of API integrations or companies who have 45,000 developers and are trying to build out the Salesforce.com PaaS strategy.
This thinking is becoming a lot more widely spread than I thought two years ago. I have talked to you about it and I have seen this coming and it is picking up. What is your analysis of this trend? How does your fund see this trend and how do you want to play in it?
Eric Benhamou: You are completely right in describing this trend as an exciting trend. It is very important, and it reflects the fact that today, you have to optimize stacks to roll out applications that make a meaningful difference in the life of consumers of enterprises.
The way to manage stack optimization is to have a platform-like approach with some rich toolsets. This is completely within the scope of what we do. An example of this would be Platform.sh. It’s a French-born company, but it has offices in half a dozen different countries. They help companies optimize and manage fleets of websites.
Sometimes we make a mistake in thinking that a website is a unique thing, but in reality, large enterprises have hundreds, if not thousands of websites that cater to different customer groups. Managing those websites requires unique toolsets. You cannot manage these websites one at a time.
The platform does this in a very effective way for extremely large enterprises. They do this by optimizing a stack that is unique to the way you build and run websites. That is an example of a capitalization on the PaaS trend.
This segment is part 2 in the series : 1Mby1M Virtual Accelerator Investor Forum: With Eric Benhamou, Founder and General Partner at Benhamou Global Ventures 2021
1 2 3 4 5