Is LEXISTEMS alone in this market? In other words, what is your competition?

On the one hand, there’s the basic AI techs like Rasa NLU and the rest of them like api.ai, LUIS, wit.ai, etc. The keyword here is “basic”. For most business projects, just a domain-specific intents + entities approach is not enough, or can’t be extended / capitalized upon. It is easy to understand and the web is full of feedbacks about that.

Advanced APIs like Keras or Google’s TensorFlow have remarkable capabilities (albeit with technical and operational constraints) but due to the NLU requirements, you need a fair amount of additional tech and expertise to achieve meaning-based results.

On the other hand we have the majors (Google, Facebook, Microsoft, IBM, Amazon…) whose public-facing technologies are mostly not for sale and mostly not applicable to private/business data, as I said earlier. One notable exception is IBM Watson. But according to some of our prospects and, again, the media, the feedback is mixed, especially in use cases with languages other than English.

Last, our main competition comes from the Academia, including the University labs that we work or interact with. But the problem here is code property and licenses. It’s a nightmare. As though complete technical integration and levelling wasn’t difficult enough, it opens the door to endless litigation. And as you know, litigation in tech boils down to legal spending power, regardless of who’s right and who’s infringing.

Another problem in this respect is the granularity of technologies. When it comes to meaning-based data processing, there’s a bit of assembly required. We know what’s going on in many labs, we do follow a number of projects and research, sometimes actively. That makes us confident that the spectrum of technologies and expertise needed to reach the level of maturity we have in this demanding field means stronger and stronger barriers to entry.

Bottom line, the competition for LEXISTEMS’ meaning-based technologies is scarce. When customer needs are limited, our competition is usually cheaper (but not necessarily faster to deploy in production). When the needs are more complex, I dare say we compare very favorably to any competitors we know of.

As a startup, what is LEXISTEMS’s stage with regard to funding?

People we talk with call us atypical in this respect – which is fine with us. What we develop costs serious money but, technically, we’re “pre-seed” since we did not seek seed funding initially. So far, LEXISTEMS has been financed by its founders – Frédéric Milliot and I – as well as by prizes from competitions we won and institutional or commercial programs we were selected in. I won’t go into details here but some of the amounts we’re talking about were pretty sizeable – in the 6-figures range.

Anyway, the company profile today is that LEXISTEMS is 100% owned by its founders, has no debt and owns 100% of its technologies, intellectual properties and trade/service marks.

Remember also that although our technologies are based on years of preparatory R&D, LEXISTEMS has been incorporated just about a year ago. The plan was to take these twelve months to ready our APIs for market and fine tune our marketing strategy iteratively through customer meetings. We’re now finishing the design of our communication in order to get public and business exposure in Europe and the Americas.

That puts us in a good position for an ambitious “Early stage” funding, which is precisely what we are going to pursue starting Q3 2017. This funding will come in addition to our first recurring revenues, as some of the customers POCs I mentioned will probably have converted by then. We’ll use the cash to further secure our technological advance and develop sales in selected high-potential European and American markets. Which, I guess, isn’t that atypical, is it?