The Inside Story of Talla’s $8.3M Series A

You may have heard today that Glasswing Ventures led Talla’s Series A, to the tune of $8.3M. I’d also like to welcome new investors PJC, Pillar, and Launch into the round, and thank Avalon Ventures, who led our seed, for their participation. It’s a huge validation point for Talla, so I want to take a few minutes and talk about why we chose the investors we chose, and what this means for our customers, and for the A.I. and “bot” ecosystems more generally.

We started Talla 18 months ago and raised a large seed round and set off to build an enterprise A.I. product. The general insight we had was that chat was going to be a thing, which meant we would now have lots of natural language data sets about work, and with new machine learning approaches to NLP, those data sets would be computable.

Most of that turned out to be true. Chat is definitely a big trend, proven out by Slack’s massive growth, Microsoft’s entry into the space, and Google’s announcement that hangouts will become a chat platform that supports bots.

The usage of bots in the enterprise is growing exponentially. We know this because of our own customer base. Our largest pilot to date is with a Fortune 50 company with over 90,000 users on Microsoft Teams. We’ve seen many Fortune Global 2000 companies with open budgets and projects for bots at work. While the consumer bot experiences may be dragging down the perception of bots, they are getting real traction in the enterprise.

On the NLP side, well, that technology has been hard. After trying almost everything we could find off the shelf, we ended up writing much of it ourselves. The secret to NLP at this point in time is actually on the UX side. We’ve learned that the more you can subtly coach the user how to respond to your bot, or set rails to keep them in a certain frame of mind, the easier it is to parse the NLP because you can assume a lot about intent. At this point, we’ve gotten pretty good at it.

Nonetheless, Series As are hard to raise. When we set out to do it earlier this year, we found out VCs could be lumped into three broad categories:

  1. The “we don’t understand this so we are going to sit back and watch” category.
  2. The “oh, this is like vertical SaaS so, you need 100K+ in MRR to get a Series A” category.
  3. The “AI is the new wave of disruption. We recognize execution-minded founders with significant IP, and the huge potential AI will have on the enterprise — and we will take a lead position,” category.

At this point, I would estimate about 25% of VC firms are in category 3. We focused there.

We had a couple of other requirements as well. First of all, I’d like to IPO this company. I believe we are onto something big, and while we are still in the early days of this market we made sure to align ourselves with investors who have vision, conviction and vast experience of investing in transformational startups that have become market leading companies. Secondly, despite having HQ in Boston, we have a small Palo Alto office and I spend a lot of time there, and we wanted either a West Coast lead or one of the few East Coast leads with a West Coast mindset — e.g. bigger rounds are good, aggressive product strategy is good, founder and team friendly companies are good.

We found that mindset and more in Rudina Seseri from Glasswing Ventures. Glasswing is a brand new fund, entirely focused on enterprise artificial intelligence. I knew Rudina socially from her time at a previous VC firm, and so we started talking last year as her fund was coming together. I’m really excited to be part of their portfolio because with such investors I can focus and execute a long term strategy of building a large, publically traded company. As a bonus, Rudina’s diligence process resulted in several potential customers for Talla, which is exemplary of the kind of investors we have in Glasswing Ventures.

Also coming into the round we added PJC, who I really liked because they do a lot of deals on both coasts, and have a great set of connections to help us as we build Talla. They also like big bets with big upside. I know that sounds like something all VCs should want but, I’ve been doing this long enough to know that being a VC can be a cushy job if you want it to be, and a lot of them focus on making safer bets that keep their gravy trains rolling vs. leading the next transformational wave and being value-added investors in the true sense of the word.

And finally, Pillar and Launch Capital came into the round as well. Adding them was a no-brainer because both Jamie Goldstein from Pillar and Woody Benson from Launch are the kind of people who say what they think, and don’t beat around the bush, but are also very supportive of their companies.

My test for a major investor is this — I know bad things will happen. If I call you and say “we just lost our biggest customer,” do we have a productive conversation or an unproductive one? Because I know that in coming years, I will have to make several depressing phone calls. It’s the nature of startups.

So I’m really excited to close the meaningful Series A round with investors who align with our founder and startup mindset, and share our vision for the future of the company. Our investors share our philosophy from not requiring non-competes (and, in fact, being supporters of doing away with them completely) to helping us with building the company. With this investor base and strong capital at hand, we can drive the growth of the company. And of course we are hiring!

As someone who runs an A.I. company and also does a fair amount of angel investing in the space, I can tell you that many many A.I. companies are struggling with their go to market strategies. As a company that is honing in on our go to market execution, I hope that the few of us that are raising larger and later rounds can start to pave the way for the ones who come after, by making investors and customers more comfortable with A.I., and helping establish a set of best practices.

For bots in particular, I think Talla’s Series A is a great validation point. And I can tell you from our funnel that chat adoption in organizations is growing very, very fast. Businesses want bots. I think over the next year you will see a lot more business focused bots land Series As.

And finally, if you are looking for a bot for an internal service team like HR or I.T., you should check out Talla. Or, read more on our blog about what this round means for our customers.