Legit

MATT OSMAN: My name's Matt Osman. I am the CEO and co-founder of Legit. Legit as a platform built for R&D teams, by people who have both been on, and led, R&D teams. So our chief scientist used to lead an R&D team. My CTO used to be on an R&D team. And the tech that we've built is redesigned to massively increase the efficiency of the R&D process at the team level.

So we use natural language processing, so a subset of artificial intelligence, to take a description of a new product, an idea, within a corporate R&D context, and in real-time our algorithms will extract all of the concepts in that new idea and match that against 30 million pieces of technical literature in real-time. And it'll tell you how similar to technical ideas are.

So instead of going down a path of trying to invent something that's already been done by the competition and wasting a bunch of money and time, we'll tell you upfront, no, there's a blocker in the way there. And then the tech that we've also built allows the engineer, the researcher, the R&D team to refine their ideas based on the feedback that our algorithms give them to become more novel. So we're trying to help R&D teams find what's new and valuable much, much quicker, using AI.

My background is I trained as an attorney, and then I went to go and work in asset management trading very exotic financial instruments. And I became fascinated by the effect that I thought artificial intelligence was going to have on a whole bunch of spaces, but particularly professional services. And I became very interested in structured data sets like legal documents and what artificial intelligence could do using that as a data source.

I came across at my desk some research out of CSAIL, it was actually my co-founders' research, it was his graduate work. He built a tax attorney, essentially, out of these things called coevolutionary algorithms. And I thought it was so unbelievably cool I stepped out of my office during my lunch break and I cold called him.

And he got this garbled voicemail message from a British person saying, hi, you don't know who I am, but I think your research is awesome and I think we should start a company. And so it so began a little bit too short of a wooing process, actually. But then Jacob Rosen, who's my co-founding CTO, who was at CSAIL, he flew over to London. We spent 10 days together in pretty close quarters to work out whether we were compatible as co-founders.

And then we added on Antony Bucci, who is our chief scientist, a little bit after that, to work on the natural language processing that we use. So it all started with a cold call, I suppose.

[MUSIC PLAYING]

MATT OSMAN: My names Matt Osman. I am the CEO and co-founder of Legit. Legit is a platform built for R&D teams by people who have both been on, and led, R&D teams. So our chief scientist used to lead an R&D team. My CTO used to be on an R&D team. And the tech that we built is redesigned to massively increase the efficiency of the R&D process at the team level.

So we use natural language processing-- so a subset of artificial intelligence, to take a description of a new product, an idea, within a corporate R&D context. And in real time, our algorithms will extract all of the concepts in that new idea, and match that against 30 million pieces of technical literature. In real time. And it'll tell you how similar two technical ideas are.

So, instead of going down a path of trying to invent something that's already being done by the competition and wasting a bunch of money and time, we'll tell you upfront, no, there's a blocker in the way there. And then the tech that we've also built allows the engineer, the researcher, the R&D team, to refine their ideas based on the feedback that our algorithms give them to become more novel. So we're trying to help R&D teams find what's new and valuable much, much quicker using AI.

My background is I trained as an attorney. And then I went to go and work in asset management, trading very exotic financial instruments. And I became fascinated by the effect that I thought artificial intelligence was going to have on a whole bunch of spaces, but particularly professional services. And I became very interested in structured data sets, like legal documents, and what artificial intelligence could do using that as a data source.

I came across, at my desk, some research out of CSAIL. It was actually my co-founder's research. It was his graduate work. And he built a tax attorney, essentially, out of these things called co-evolutionary algorithms. And I thought it was so unbelievably cool. I stepped out of my office during my lunch break, and I cold called him. And he got this [LAUGHS] garbled voicemail message from a British person saying, hi. You don't know who I am. But I think your research is awesome, and I think we should start a company.

And so began a fairly-- little bit too short of a kind of a wooing process, actually. But then Jacob Rosen, who's my co-founder and CTO, who was at CSAIL-- he flew over to London. We spent 10 days together in pretty close quarters to work out whether we were compatible as co-founders. And then we added on Anthony Bucci, who is our chief scientist, a little bit after that to work on the natural language processing that we use. So it all started with a cold call, I suppose.

[MUSIC PLAYING]

MATT OSMAN: In terms of what we look for in partners, there is no particular industry that we would necessarily favor from a technical standpoint. The technology we've built is meant to be industry agnostic, and we think we can provide value to pretty much every industry and also every geography as well.

The profile of the partners that I think we have worked best with in the past, it's highly collaborative. It requires a lot of active participation from the corporate partner. So they need to have a pretty innovative mindset, so those are the relationships that have kind of work best.

Stanley Black & Decker is obviously a company that we're working with. We're working with our breakthrough innovation team that they have here in Boston, and that is an example of a very, very collaborative partnership that we've had. But then we also work right the way down in terms of scale to a very early stage medical device companies. We're working with like an eye implant company at the moment with a very small engineering team all the way up to potentially a very, very large deployment with a life sciences company, which would be in the order of thousands of users.

So it kind of runs the gamut, which is nice for us I suppose. We recently raised $2.6 million from a mixture of venture capitalists, many in the West Coast and then some Boston Angels. The Boston Angels in particular are very focused on the artificial intelligence ecosystem, which is incredibly strong in Boston and has been historically.

So that money is going to be used to scale up our sales team and our engineering team, so we'll be growing headcount over the next year or so, I think, to about 20. We may need to get a new office, so it might go in new office space as well. Kind of rarely for a company our size, we have an R&D department. It's one of the reasons that I think that we're such an interesting partner for large corporates is because a lot of the pain they're feeling, we feel.

We have research scientists whose job it is to do everything from basic research on our NLP through to kind of moonshot stuff, wouldn't it be great if we can read the mind of an engineer, and then everything in between. I think it's one of the things that makes us such a good partner to work with.

And so we'll be scaling up on the R&D side as well and also launching a couple more products this year. So one of the things that we're going to add with the money is the ability to identify people based on ideas. So as a user at one of our corporate R&D clients, who is working on a new idea, for example, we will match that against a bunch of technical literature and identify if there's any one external to the company that is working on something similar and try and broker an introduction, especially if that person isn't academia. So that's a very, very exciting thing. So sales engineering, I think, and some more R&D.

We've made a very conscious decision about diversity is as a hiring policy. I think that given the scope and the scale of the company that we're trying to build, having representation of different groups early is incredibly important. I also think that I'm incredibly involved in hiring right now, and I'm already stretched pretty thin. Your company will look like your first 10 employees, and culture and diversity pays dividends not just in the short term, but even more in the medium and long term, and it's something you need to get right very, very early, because there becomes a tipping point where the culture can become toxic, and it becomes harder to hire and harder to do great work, because actually, yes, we have intellectual property, and we have a bunch of really cool tech, but actually the thing that brings value to the people that we work with is the quality of our team.

MATT OSMAN: You know, one of the key things that we've learned about innovation through this process is testing repeatedly and early. The technology use case that we initially envisioned when I first reached out to Jacob has changed slightly, and that came through market feedback. So you have to be very, very open to kind of iterating quickly.

There's the Allen Ginsberg quote about killing your darlings, which I think is pretty apt when it comes to ideas. And I'm wrong most of the time, but I know it, which I think is the crucial point. And so that is that is a really, really important thing about innovation.

The other thing that I think that we've learned, especially because we have an R&D department is that R&D is often-- it's a really hard thing to do innovation. So one, you can't buy it. You can buy the ingredients of innovation. And that's broadly what we're trying to do. We're trying to sell the ingredients of innovation.

First we're going to sell software applications to a large corporate R&D departments. Then we're going to try and broker introductions to the right talent. Eventually, we'll start brokering introductions to the right suppliers and materials until we become kind of a one stop shop for the ingredients of innovation. So that's one of the most important things that we've learned.

We have a series of large deployments currently with big enterprise clients. So we'll be focusing on making sure that both parties get the most amount of value out of that. But we're always taking on new partners, and we're growing very rapidly. The team over the next year, I think will grow to, I think, probably over 20.

We have a pretty extraordinary product and feature launch cycle. We launched a new product over-- took us about a month to launch a new product. We launch a couple of features every two weeks, so we move very, very, very rapidly. The expertise locators are finding people that are similar to your idea as something that's going to be coming in the next few months.

And then I imagine at some point we will go out for another round of fundraising to accelerate our growth and potentially move beyond the US market as well. Yeah, so the are our current products is called Legit Invent. And it's an application for engineers, researchers, and scientists to identify whether what they're working on is new in comparison to other inventions, pieces of technical literature.

So we use natural language processing to identify how similar two pieces of technical literature are, so your idea and then what's gone beforehand. It will also start to identify other people within your organization, who have similar ideas or similar talents to try and create kind of, you know, Navy SEAL teams to really tackle hard problems to identify those collaboration opportunities early, and also identify potential partners outside of the corporation that you work for. But it's all based on the comparison of two pieces of technical text of any length really.

Hi, my name's Matt Osman. I'm the CEO and co-founder of LEGIT. We're an AI company that is focused on improving R&D efficiency. We were built by members of R&D teams, for R&D teams, and we originated out of MIT's Computer Science and AI lab.

The problem we're trying to solve is this, there are ever increasing costs of R&D, and there's increasing pressure on research and development activities to drive top line revenue. But it's even harder, because of the morass of technical literature, to identify whether new ideas and products are in fact ahead of the competition or whether they're new. And assessing this competition is a time consuming and highly laborious process.

Our natural language processing model learns through user feedback and is trained specifically on the usage patterns of the clients that we work with, building a scalable and valuable model of the inventive landscape that is particular to ever single company. Using LEGIT allows you to massively increase research engineer and scientist efficiency. It reduces your time to market by streamlining the ideation process.

Companies like Stanley Black & Decker are using us already to reduce R&D waste. And it also provides automated and deep insights on the competition, and allows you to find collaborators, partners, and potential hires based on your R&D initiatives. Thanks very much.

[MUSIC PLAYING]