Littlebird raises $11M for its AI-assisted ‘recall’ tool that reads your computer screen

There has been a batch of speech astir gathering context for AI systems. In user software, we person seen startups being built astir search, documents, and meetings. All of them privation to seizure discourse from your integer life, supply connections to different tools, and fto you query each that data. Some tools went further. For instance, Rewind (which became Limitless and sold to Meta) and Microsoft Recall purpose to seizure everything happening connected your surface and assistance you retrieve it all.

A caller startup called Littlebird is trying a akin happening with a somewhat antithetic approach. While apps similar Rewind store screenshots oregon immoderate benignant of ocular data, Littlebird is “reading” the surface and storing the discourse successful substance format.

The halfway thought down the merchandise is that since it is speechmaking your surface each the time, you don’t request to supply further discourse for productivity. The startup believes that portion a batch of AI tools are trying to distract you, Littlebird tin enactment successful the inheritance and tin lone look erstwhile you privation it to.

Image Credits: LittlebirdImage Credits:Littlebird

When you acceptable up Littlebird connected your computer, you tin customize which apps you privation the app to disregard and not seizure immoderate context. The startup said that it automatically ignores password managers and delicate fields successful web forms similar passwords and recognition paper details. You tin opt to link different apps similar Gmail, Google Calendar, Apple Calendar, and Reminders with the app, arsenic well.

The app lets you inquire questions astir your data, offering pre-generated prompts to get you started, specified arsenic “What person I been doing today?” oregon “What benignant of emails are important to me?” In a mates of days of usage, I noticed that these prompts became much personalized arsenic clip went on.

Littlebird besides has an in-built Granola-like notetaker that uses strategy audio and runs successful the inheritance to seizure transcription from meetings and make notes and enactment items based connected that. When you unfastened up a gathering successful the elaborate view, there’s an enactment called “Prep for meeting” that takes the discourse of past meetings, emails, and institution past into relationship to springiness you much details astir the meeting. The diagnostic besides fetches accusation from sources similar Reddit to pass you what users are reasoning astir a peculiar merchandise oregon a company.

Image Credits: Littlebird

Another instrumentality called Routines offers elaborate prompts for Littlebird to tally astatine a repeated interval, specified arsenic daily, weekly, oregon monthly. The institution lists immoderate ready-to-use routines similar regular briefing, play enactment summary, and yesterday’s enactment summary. Users tin make their ain routines arsenic good with customized instructions.

Littlebird was founded by Alap Shah, Naman Shah, and Alexander Green successful 2024. Brothers Alap and Naman founded Sentieo, a level for organization investors, which was sold to marketplace quality steadfast AlphaSense. They antecedently besides co-founded a steadfast nutrient institution called Thistle. Alap was besides a co-author of the viral Citrini insubstantial connected however AI agents could destruct the economy, which resulted successful assorted tech stocks dipping. Green has built assorted companies successful hardware, software, and AI.

“We got started erstwhile Alap posed an absorbing occupation that AI is going to beryllium astir your [users’] data. Models don’t cognize thing astir you, and that limits their utility. We were reasoning astir assorted UI and OS paradigms that were apt to beryllium ripe for disruption with AI and that kicked disconnected Littlebird arsenic a project,” Green told TechCrunch implicit a call.

Green noted that portion Rewind was adjacent to what Littlebird is trying to do, it relied connected screenshots and didn’t person a large hunt experience. He said that the startup is conscionable getting started and determination are galore much problems to solve, including making ample connection models (LLMs) recognize antithetic kinds of discourse astir users.

With Littlebird, users tin region their information astatine immoderate time, and their information is stored successful the unreality with encryption. Green said that the rationale down storing the information successful the unreality was to tally almighty models for antithetic AI workflows, which is not imaginable locally.

“We don’t store immoderate ocular information. We lone store text, which makes the information a batch lighter-weight. I deliberation that was astir apt different crushed that Recall and Rewind struggled, which is that taking a screenshot is simply a batch much information hungry. I besides deliberation it’s much invasive,” helium said.

Image Credits: Alexander Green

Littlebird is escaped to download and use, but to get much usage limits and entree to features similar representation generation, users tin wage for plans starting from $20 per month.

The startup has raised $11 cardinal successful backing led by Lotus Studio, with information from Lenny Rachitsky, Scott Belsky, Gokul Rajaram, Justin Rosenstein, Shawn Wang, and Russ Heddleston.

Several of these investors are regular users of the product. Rajaram, who has worked astatine Google and Facebook connected advertisement products, said that the merchandise removes the friction of remembering, retrieving, and re-explaining your ain work. DocSend co-founder and CEO Heddleston said that helium rewrote the company’s selling tract utilizing the tool, utilizing discourse from meetings, email, Notion, and more.

Rachitsky, who runs his ain newsletter and podcast, said that AI is arsenic bully arsenic the discourse it has, and it misses truthful overmuch astir your day. He said helium asks the instrumentality astir improving his productivity workflows and being happier. He said that for semipermanent success, the merchandise volition request to find a slayer usage case.

“I deliberation it’s each astir uncovering that slayer must-have usage case. That’s each that matters to this product’s occurrence close now. I cognize a batch of radical already person recovered that for themselves, and the squad is leaning into these experiences arsenic they spot these usage cases emerge,” helium noted.

“I’ve had a batch of AI merchandise builders connected the podcast, and the astir accordant taxable is that you don’t really cognize however radical volition usage your merchandise until you enactment it out. The strategy is to enactment retired aboriginal stuff, spot however radical usage it, and treble down connected those usage cases versus waiting for thing wholly figured out.”