Microsoft’s huge purchase of health tech AI company Nuance led the technology news cycle this week. The $19.7 billion transaction is Microsoft’s second-largest to date, only beaten by its purchase of LinkedIn some years ago.
For the AI space, the sale is a coup. Nuance was already a public company, but to see Microsoft offer a firm premium over its public-market value demonstrates the value that AI technology can have to wealthy companies. For startups working in the AI space, the Nuance deal is good news; the value of AI revenue was repriced by the acquisition’s announcement — and for the better.
In light of the megadeal, The Exchange dug into the AI venture capital market. What’s happening on the startup side of the coin in the artificial intelligence and machine learning (AI/ML) space?
The Exchange explores startups, markets and money. Read it every morning on Extra Crunch, or get The Exchange newsletter every Saturday.
To get a handle on the situation, we’ve compiled Q1 2021 and historical venture capital investment data via PitchBook, spoken to an active venture capitalist with a focus on AI-powered startups, and heard from a couple of startups recently featured on CB Insights’ list of leading AI upstarts for their take on the recent news.
The picture that emerges is one of strong investor interest and the expectation of even more in the wake of the Microsoft-Nuance tie-up. For AI startups, it’s a great time to be in the market.
This morning, we’ll start with a look into recent venture capital activity in the AI/ML market and its historical context. Then we’ll talk to Zetta Ventures’ Jocelyn Goldfein and a few companies in the AI space. Let’s go!
A venture capital rush
According to historical data compiled by PitchBook, venture capital investment into U.S.-based, AI-focused startups is enjoying a strong start to the year. Per the group’s provided dataset, from the start of 2021 through April 12, or the first 101 days of the year, 442 deals in the space were worth $11.65 billion.
In 2020, the same query for U.S.-based startups working in the AI and ML space — the line between ML and AI is blurrier than ever — turned up 1,601 rounds worth $27.49 billion.
In terms of round volume, then, 2021 does not look particularly special. However, AI and ML startups in the U.S. will smash 2020’s venture capital totals if they maintain their current pace. Doing some quick run-rate calculations, if AI and ML startups kept up their year-to-date pace, they would raise 1,597 rounds worth a total of $42.1 billion.
That dollar amount would be the strongest in history, easily eclipsing 2020’s prior record of around $27.5 billion.
What’s driving the huge boom in dollars invested into U.S.-based AI and ML startups? As The Exchange has examined several times in other sectors, it’s the late-stage money that is shaking up the market. Through April 12, 2021, PitchBook counts $9.08 billion in late-stage money going into our segment of startups. That’s just about half of 2020’s total for the same slice of the same market, and only fractionally lower than full-year, late-stage totals in both 2018 and 2019.
Early-stage and seed data look strong so far, but as those two categories of the venture capital market have the most temporal lag in their reporting, we don’t want to draw too many conclusions yet. The same reason is why we’re not shouting about a possible, if modest, decline in 2021 AI and ML startup round volume in the United States.
Sure, our extrapolated run-rate figure for this year’s expected AI and ML startup rounds in the U.S. is under 2020’s total, but we expect more early-2021 investments to become known over time, boosting the tally. The dollar figure is more sturdy at this point in time, so we focused our attention there.
The apparent boom in funding is not lost on the venture capital set. AI-focused Zetta Ventures’ Jocelyn Goldfein — TechCrunch has spoken with her before — told The Exchange in an email that both the last quarter of 2020 and the first quarter of 2021 were “great quarters for AI entrepreneurs,” with the venture capitalist going on to praise both the “quantity and quality of entrepreneurs raising seed rounds” in the space.
But it wasn’t only the earliest stage of investors who were enthused by the prospect of putting capital to work in AI/ML startups. Per Goldfein, her firm “also saw tremendous enthusiasm from Series A and B investors, especially around the tooling and infrastructure for AI projects (MLops and data infrastructure.)”
In collating information and notes for this particular entry, The Exchange read through CB Insights’ AI 100 list, which included a few trends that the group felt were worth highlighting. Goldfein said that her firm’s “taxonomy is a bit different,” citing themes like “data quality, model quality, active learning, annotation, synthetic data,” along with what she described as “major horizontal markets” like the public cloud and “mega verticals” like financial technology and health tech.
Let’s dig into a few of the trends that Zetta and CB Insights both noted to get a better grip on what’s going on inside the hot AI/ML startup market.
Inside the AI/ML startup market
One of the key trends that becomes obvious after even a quick look at CB Insights’ 2021 AI 100 picks is that healthcare is hot. As the write-up points out, “a number of these companies developed new products and features directly in response to the pandemic to mitigate its impact and help clients adapt.” But it is not just that: Healthcare is the most represented category among featured startups that focus on a specific industry.
This also matches a recent comment from Microsoft CEO Satya Nadella that caught our attention in the middle of the Nuance acquisition announcement: “AI is technology’s most important priority, and healthcare is its most urgent application,” he said. This is representative of how our sources view the Nuance deal: as reflecting excitement around the voice space, but also around its applications, and, more broadly, around AI.
Startups that work directly on voice technology, speech recognition and natural language processing (NLP) are likely to benefit from being in Nuance’s wake. In the words of Deepgram‘s CEO, Scott Stephenson, “Microsoft’s acquisition of Nuance will intensify VC’s interest in the speech market the same way that GM’s acquisition of Cruise did in the self-driving car market. There will be an immense inflow of cash to fund speech companies now that VCs know that investments in speech pay off.”
As you may recall, Deepgram’s offering centers on automatic speech recognition (ASR), for which they leverage AI and deep learning in order to help its users build voice-enabled applications. This has proven to be a good bet for the company itself: After raising a $12 million Series A round in March 2020, it attracted Tiger to lead its $25 million Series B earlier this year — and is now part of CB Insights’s AI 100 shortlist.
Looking at the bigger picture, Deepgram’s momentum reflects the broader trend to which it is contributing: Voice technology is no longer reserved for tech giants. “AI has made custom, tailored speech recognition not only possible, but also affordable,” Stephenson told us. As a result, it is being applied across a wide range of industries.
Edtech startup and fellow 2021 AI 100 winner Elsa is a good example of this because it is using speech recognition technology to help ESL speakers correct their pronunciation (see our recent post). CEO Vu Van noted: “Speech recognition technology has been gaining more and more attention from the VC world given how companies leverage voice recognition technology on multiple fronts these days — in healthcare, education, productivity in the workplace, fintech and more.”
That optimism could spur even more venture deals, and even more big sales in the AI/ML startup space. Stephenson called the Nuance deal “evidence of a decades-long AI market expansion,” adding that the transaction is “just the tip of the iceberg.”
In another few months, we’ll have enough data to better forecast the year’s venture appetite for AI/ML deals more precisely. But from what we can see thus far, it’s going to be a scorcher of a year.