This former Facebook product manager started his first company at 12. Now, he’s trying to take on PowerPoint with his $300 million AI startup.

Welcome to this special edition of LinkedIn News Tech Stack, which brings you news, insights and trends involving the founders, investors and companies on the cutting edge of technology. It’s created by Tech Editor Tanya Dua and you can check out our previous editions here .?

First: What are your tech predictions for 2024? We want to hear your biggest, boldest ideas. Fill out this short survey to share your opinion: https://lnkd.in/ewW5BG9i

Pitch me the interesting investors, founders, ideas and companies powering emerging technologies like AI. Share your feedback and follow me on LinkedIn for other tech updates.

In some ways, Keith Peiris was destined to be a founder.

The former Facebook and Instagram product manager started his first company at the age of 12, when his dad turned his side hustle of building interactive websites at the peak of the dot-com craze into a full-fledged family business .

Two decades later, Peiris is riding yet another wave. The 35-year-old is now the cofounder and CEO of buzzy AI startup Tome , which taps into artificial intelligence technology to let people create presentations spanning text, images, audio and video in just a few clicks.

“I've worked on communication products my whole life,” Peiris said. "We saw a gap in the market where visual storytelling tools were abundant, but tools for sharing complex ideas and concepts were lacking – that's why Tome was born."

The San Francisco-based company is one of 13 U.S. upstart businesses using AI in meaningful ways on LinkedIn’s Top Startups 2023 list, a data-driven ranking of the young companies that are growing fast and gaining attention from job seekers and investors alike.

Tome wants to pioneer a new storytelling format

It’s not hard to see why AI startups feature so prominently on this year’s list.

Ever since the launch of ChatGPT last November, the technology has captured the collective consciousness. It’s been billed as holding the same transformative power as the computer, the internet and the mobile phone, and it’s reinventing business functions across the board . AI startups have also been a rare bright spot in an otherwise gloomy year for the overall venture ecosystem – with AI’s share of startup funding in the U.S. doubling in 2023, even as overall funding fell by 50%.

Tome is a beneficiary of these trends. The company has raised $81 million in funding so far (and is reportedly fielding more interest ) from blue-chip investors including Greylock , Coatue and Lightspeed at a valuation of $300 million. It also boasts a rapidly growing user base, claiming to have surpassed the milestone of 10 million users in its first nine months. Meanwhile, it has expanded its headcount from 45 to nearly 70 employees over the last year, including the hire of machine learning leader Ves Stoyanov , who spent nearly a decade at Meta .

Peiris and cofounder Henri Liriani – another former Meta product manager – are betting that recent advances in AI can help them pioneer a new storytelling format. In the way that ordinary language takes vague or abstract thoughts and makes them concrete, Tome says it aims to streamline professional communication and productivity, using AI to drive an idea from concept to creation in quick succession.

“We’ve made it so easy to share what you're eating, what you're doing with your family and where you're going on the weekends, but if you want to share something in your head – like an idea or a concept – you have to use PowerPoint or Medium,” Peiris said. “We want to take what we did at Instagram and what others did at Snap and TikTok with storytelling into ideas and concepts.”

Their investors are enthusiastic about the vision.

“Tome was designed to use different media types and integrate data from multiple sources, and AI lends itself really well to that as it helps generate those images and texts,” said Greylock general partner Seth Rosenberg , who helped incubate the startup alongside investor Reid Hoffman , LinkedIn ’s cofounder.

For Lightspeed’s Michael Mignano , who led Tome’s most recent funding round, whether it’s text, photo or video, all forms of media are ultimately democratized by technology.

“Some of the most iconic companies in the world are the ones that take advantage of this democratization and come up with their own proprietary formats,” he said.

The company is using AI to reimagine how traditional presentations are made

So, how exactly does it work?

Tome is essentially an AI-powered canvas powered by large language models including GPT4 as well as its own proprietary LLMs derived from open-source AI models, so it can take a prompt and spit out everything from a resume to a deck within seconds. Its software allows users to select their output – such as a page or presentation – and set parameters like length. Then the AI models jump into action to populate the canvas with images and texts based on the prompt. Users can also remix the results by asking Tome to generate something again until they’re satisfied with it.

Tome reimagines the slides found in PowerPoint or Google Slides as dynamic tiles, with changes to one automatically updating the others. The product also comes pre-loaded with multiple templates and prompts, like “Create a presentation about” or “Generate an image of.”

Los Angeles-based interior plant designer Zilah Drahn said that she always struggled with ways to pitch her small business, “Plants & Spaces,” given its niche focus on plants and because she never quite had a way with words. Tome helped her create everything from project decks to digital care guides for clients, and she said she’s now booked out through the rest of the year, while also charging clients more per project.

“Tome has been a lifesaver,” she said. “It’s made us look elevated, saved us a lot of time and enabled us to charge more.”

Luke Thompson , chief operating officer at ActionVFX , which creates visual effects and stock footage for TV shows and films, is another user. Tome offered an instant opportunity for the studio to improve its pitch presentations to clients, he said, since it lets users easily embed prototypes of videos and graphics, which were historically embeds of static screenshots.

Tome faces significant headwinds from incumbents

The ongoing AI gold rush is an obvious tailwind, but Rosenberg’s conviction in Tome also rides on its cofounders and their previous experience in building and scaling products. Both of them worked on products like Instagram and Facebook Messenger, which are used by billions of people. (Rosenberg himself used to work at Meta).

“One of Keith’s superpowers is eliciting the creativity of others, and Henri is one of the most creative people in the world,” said Rosenberg. “Henri has gone a long way from turning Keith’s problem statement and understanding of the customer and the market into a real product.”

But converting initial traction into a sustainable revenue stream won’t be easy.

After officially launching its product last September, Tome rolled out a paid version priced from $8 to $10 a month for individuals and teams this April, and also offers enterprise pricing. And while it claims to have a base spanning founders, sales professionals, marketers and product managers at small businesses and companies including Snap Inc. , Meta and 三星电子 , most of its users take advantage of its free version, which has spread largely through word of mouth.

The company declined to share any revenue metrics, though The Information recently reported that it has passed $2 million in annual run rate.

A company spokeswoman highlighted that it’s been less than six months since Tome launched its first paid tier. But experts said Tome must boost sales and land enterprise deals if it wants to capitalize on its momentum and chip away at the stronghold held by incumbents.

“Users might be an early venture metric in a lot of ways, but startups have to start to build business models that actually drive revenue,” said George Mathew , managing director at venture firm Insight Partners , which isn’t an investor in Tome. “That's the lens that every business is measured by over time.”

Another investor, who considered funding the company but ultimately didn't, told LinkedIn News that they thought it was better to hold off given Tome was “still very early in terms of revenue.”

Taking on deep-pocketed rivals like 微软 and 谷歌 is no small feat. Not only are their Office 365 and Workspace suites multi-billion-dollar businesses used by enterprises globally, but they are also making significant AI investments of their own. Microsoft (LinkedIn’s parent) announced its unified AI assistant , Copilot, across products from PowerPoint to Teams last week. Other competitors, like Canva , Coda and Notion , have announced AI features of their own.

And, as has been the case for companies like Snap – which pioneered the Stories format on social media that was quickly and widely copied by others – being a pioneer isn’t necessarily a defensible advantage.

“I don't know that there's much in Tome that Microsoft Copilot can't do more exhaustively and better,” said J. P. Gownder , vice president and principal analyst at Forrester covering the future of work. “That is a big problem because these are the applications that people already know and use. Microsoft and Google are continuing to invest and make their products better at a rapid clip.”

The company is making a bet on enterprises to propel it forward

To Peiris, the very things that experts point to as Tome’s disadvantages are its strengths.

“If you’re Microsoft,” he said, “you probably can't redesign all of PowerPoint for natural language – because you have all of these enterprise users that will say, ‘Hey, don't change the thing that I grew up with.’”

Still, Tome’s leaders know that they must continue to innovate – and fast. To do so, they plan to continue to beef up on engineering and product talent. They’re also taking insights gleaned from Tome’s paid tiers to shape their product roadmap as well as double down on features that would make the product more enterprise-friendly.

They’re working, for example, on integrations with other enterprise tools such as Zendesk, Salesforce and Looker . The team is also taking signals from how users interact with its platform to train and fine-tune its own large language model, making the product more personalized.

“Product-led growth has always been our default, but now we're learning there might be some sales-led growth available,” Peiris said. “In order to get there, we're going to have to do a lot of data source ingestion to be useful for people at work.”

While it started off with a mission to disrupt storytelling, Tome has expanded its ambitions by setting its sights on conquering the productivity tools market, à la Slack or Canva – perhaps with some pushing from its investors.

“Productivity is a multi-hundred-billion-dollar market,” said Greylock’s Rosenberg. “There have been enough new technology waves with mobile, the cloud and now AI, where there's now an opening to build a new productivity suite from the ground up.”

Not a week goes by without a flurry of AI-related news. This week was no different. Here’s what you need to know:

• ChatGPT can now ‘talk’ and ‘listen,’ with OpenAI unveiling a new version that allows users to speak aloud or upload an image to prompt its AI chatbot. The new ChatGPT features are slated to be available with a paid subscription within two weeks. OpenAI is also looking to roughly triple its valuation in less than a year to as much as $90 billion, The Wall Street Journal reported , citing unnamed sources.
• Relatedly, OpenAI’s new voice technology has larger implications for companies like Spotify , which announced it is working with OpenAI to translate podcasts into different languages using synthetic versions of hosts' voices. Separately, Getty Images announced that it was launching a “commercially safe” generative AI tool for images .
• That’s not all. OpenAI CEO Sam Altman, iPhone designer Jony Ive and ソフトバンク CEO Masayoshi Son have been in discussions about building a new AI hardware device , The Information reported.
• Meanwhile, Amazon plans to invest as much as $4 billion into OpenAI competitor Anthropic as part of a deal where Anthropic will use Amazon Web Services as its primary cloud provider and Amazon-made chips in its AI software.
• Meta, on the other hand, announced a series of AI updates at its Meta Connect event this week, including an “advanced conversational assistant” on WhatsApp, Messenger and Instagram in beta; multiple AI personas; and AI Studio , which will let companies build chatbots for use on Meta platforms.
• And lastly, YouTube is tapping into AI to fend off TikTok. AI will power several newly announced tools on the Google-owned video platform, including one that allows users to add AI-generated videos or photos in the background of short-form videos. Users will also be able to use AI to generate personalized ideas for new content.

• The FTC came out with its highly anticipated lawsuit against Amazon , alleging that the e-commerce giant is an illegal monopoly that has squeezed merchants and favored its own services.
• Catch up on the latest edition of our weekly Q&A series, VC Wednesdays, where investor Evan Nisselson lays out why the future of AI is visual.

Thanks for reading. Please share Tech Stack and forward it around if you like it! And if you have any news tips, find me on InMail .