EN
Back to the archive

The archive · Developer & Business Tools · Product decision · 2020–2025

Tome's AI-slides bet: 20M users and $81M, then the product died in April 2025

Ex-Meta product leads built the fastest productivity app to 1M users; when 20M users wouldn't pay, they killed it and rebuilt as Lightfield CRM.

Tome

The betThat a general-purpose AI storytelling format — a prompt becoming a whole presentation — could become a generational company while most of its 20M users stayed free.No longer exists

What the business is

Tome was an AI presentation and storytelling platform: type a prompt and it generated a scrollable, tile-based deck with text, images and layouts, with live embeds from Figma, Looker, Framer and YouTube.

Starting capital$81M across three rounds: $6.3M seed (Greylock), $26M Series A (Coatue), $43M Series B (Lightspeed, Feb 2023) at a $300M valuation

How it started

Founded in 2020 by Keith Peiris and Henri Liriani, former product leaders at Facebook and Instagram, who saw generative AI removing the friction of visual storytelling. Tome launched publicly in Sept 2022 and crossed one million users within four months — a growth record the company touted in its Series B announcement.

What happened

A $43M Series B in Feb 2023 (with angels including Eric Schmidt, Stability AI's Emad Mostaque and Adept's David Luan) followed viral consumer growth to ~10M users by mid-2023 (AI Wiki). But almost no one paid: Peiris said in April 2024 that building for millions of non-paying consumers was unsustainable, and the company cut ~12 of 59 staff to chase enterprise sales (AI Wiki).

How it ended up

On Oct 15, 2024 Tome announced it would end its presentation product; creation stopped in early 2025, features were shut down in April 2025 and user data was deleted (AI Wiki). The team shrank to a core of engineers and rebuilt as Lightfield, an AI-native CRM, which launched publicly in Nov 2025 with 100+ early customers (VentureBeat).

Background

Tome was an AI storytelling platform founded in 2020 by Keith Peiris and Henri Liriani, former product leaders at Facebook and Instagram. Its pitch was that generative AI had made the presentation tool obsolete: type a prompt and Tome generated a scrollable, tile-based deck with text, images and layouts, embedding live content from Figma, Looker, Framer and YouTube. It launched publicly in September 2022 and became the fastest productivity tool ever to cross one million users.

The growth was real but free. Tome raised $81M across three rounds — a $6.3M seed from Greylock, a $26M Series A from Coatue, and a $43M Series B from Lightspeed in February 2023 at a $300M valuation — and reached roughly 10M users by mid-2023 and 20M by 2024 (press release; AI Wiki; VentureBeat). Yet almost none of those users paid, and CEO Keith Peiris admitted in April 2024 that serving millions of non-paying consumers was unsustainable.

In October 2024 Tome announced the end of its presentation product; creation stopped in early 2025 and the product shut down in April 2025 with user data deleted (AI Wiki). The founders kept a core engineering team and rebuilt the company as Lightfield, an AI-native CRM that launched publicly in November 2025 with more than 100 early customers — a deliberate pivot away from a beloved product that was never a business (VentureBeat).

What has to be true

  • Tome is the clearest counter-example to 'growth is everything': 20M users, an $81M war chest and a $300M valuation were abandoned because the free tier could not be converted.
  • The founders made the bet explicit — that AI storytelling would become a generational productivity format — and then publicly reversed it, making the decision auditable.
  • The pivot to Lightfield shows the same team, same capital and same AI thesis can be redirected when the original wedge turns out to be a feature, not a company.
  • The full arc is documented by the company's own press release (Feb 2023), VentureBeat's November 2025 report, and a dated sunset timeline.

What can be applied

Viral consumer growth on a free tier is not a business: when the metric is users, not payers, even 20M people and $81M get killed — the founders deleted a beloved product rather than let it limp.

Aftermath

As of 2026-09-02 the Tome slides product is dead — creation stopped in early 2025, features were shut down in April 2025 and user data was deleted — while the same founders' company continues as Lightfield, an AI-native CRM that launched publicly in November 2025 with more than 100 early customers (VentureBeat). Tome's story is now a standard reference for hyper-viral AI consumer products that still were not businesses.

Sources

spotted an error? The archive wants to know.

Your turn

You just read one. Describe what you are building, and see who is betting on the same thing.

Free account · 3 free questions · no card

Related cases