Accel’s AI Investments Keep The Focus On Applications And Tooling

This article is part of our series of profiles that spotlight an investor firm’s AI strategy, including a list of its AI portfolio. In a previous interview, we talked to Sameer Dholakia from Bessemer Venture Partners, who predicted a fast adoption curve of AI through APIs. 

With 40 years of investing under its belt, global venture capital firm Accel has long seen the potential of artificial intelligence. Now the firm is making fresh bets in AI startups.

The power of AI unleashed by large language models like ChatGPT has led to significant investor interest in applying AI to everything from chips to data storage and AI infrastructure, as well as tooling and applications built on the technology.

We spoke with Accel partner Daniel Levine about the firm’s AI investment outlook.

The firm is making investments in AI across its team in India, London and the U.S., as well as its growth team. Levine likens it to investing in software, where the responsibility does not sit with one team or person but is distributed. While not every investment is necessarily in an AI company, AI is key to the firm’s strategy. Accel is unlikely to invest in chips or data centers, said Levine. But the tooling and applications side of the technology is in its wheelhouse.

“You’ll see new companies for which AI makes the difference between the product being so-so and potentially [being] a game changer in its category,” said Levine.

For many software companies, that poses a significant opportunity and threat. This technology can propel a company forward, but can also leave a company trailing as competitors take advantage of the possibilities it presents.

Ground-floor investment

Levine led Accel’s 2017 Series A funding in Scale AI, a company which trains data for AI capabilities. Scale has raised close to $600 million since then, from its Series B through Series E, which were led, respectively, by Index Ventures, Founders Fund, Tiger Global, Dragoneer and Greenoaks.

Co-founded by Alexandr Wang and Lucy Guo, Scale AI started out in this generation’s equivalent of a garage — the basement of investor Levine’s San Francisco home. The company hired thousands of contractors to annotate pictures for self-driving technology companies which today include Waymo, Lyft and Toyota. It has since expanded to natural language training as well as computer vision processing, supporting companies such as OpenAI, Pinterest and Airbnb.

Levine understood from his early days at Crunchbase — he was a colleague of ours who joined straight out of Yale — that having people looking at data under the hood is a necessary part of the process.

Application

“We’ll always play at the application level,” with the goal to make users more productive, said Levine.

Accel led the 2020 Series B funding in Toronto-based Ada, an application layer customer service chatbot that utilizes AI.

“We could assume that over time, AI can make support agents more efficient,” said Levine. These tools balance when a human or an algorithm is involved to provide a better service. “There’ll still be cases when you’re going to want to talk to a person either on chat or phone. But there are definitely a lot of cases where you want to talk to an algorithm because it’s faster, it’s more convenient and it’s consistent,” he said.

Levine referenced project management software Notion — not an Accel portfolio company — several times in our conversation. “Notion AI is a very important product for them,” he said. Notion hosts wikis, docs and projects for companies, and has launched AI tools to help with writing, summarizing and planning, to make users more effective.

Tooling

Tooling and infrastructure companies are a core AI investment focus for Accel. These are the companies that assist other companies with their tech products. One of its portfolio companies, AssemblyAI, for example, provides infrastructure around understanding audio.

Accel also announced this week it led a Series C funding in London-based generative video company Synthesia, which makes video production possible for companies without actors, cameras or studios.

Another area of interest for Levine and Accel are new database products for understanding bias or trust and safety around AI, what the model is doing, and how it drifts.

The companies that comprise Accel’s AI portfolio span more than a decade and are not only represented by the new wave of companies. The companies listed here were founded as early as 2011 and the most recent in 2021. Levine’s investments in companies that predate AI but have since integrated these capabilities include Sentry, an error-monitoring tool for software teams, and visual workplace collaboration platform Whimsical.

“You’ll see a lot of companies that in hindsight will look like they’ve replaced existing software players,” he said.

Levine has also led Accel’s investments in, and serves on the boards of EdgeDB, Gem, Mux, ReadMe, Scale, Sentry, Sprig, Vercel and Whimsical.

Crunchbase Pro Query

Accel’s AI portfolio companies

Illustration: Dom Guzman