Poke is a Palo Alto-based startup with a personal AI assistant you can reach right through messaging apps you already use—iMessage, SMS, Telegram, and in some places, WhatsApp. They launched publicly in March.
The service lets you automate daily tasks like calendar management, health tracking, smart-home controls, email alerts, photo edits, and news briefings. You do it all through simple text commands and shareable recipes.
Poke actually came out of a pivot. Early beta users started using their original email assistant for way more things, so the team adapted. Now, there’s no dedicated app and no signup process. You just go to poke.com, put in your phone number, and you’re set.
Each task gets routed to the AI model that’s best for the job. Poke mixes models from big providers with open-source options to cover a lot of ground.
How Poke works: technology, reach, and model selection
Poke fits right into your usual messaging workflow. No new interface to learn—just use the platforms you already rely on.
They use Linq to connect iMessage, SMS, Telegram, and, where possible, WhatsApp. WhatsApp’s availability is spotty, thanks to Meta’s restrictions and fees. There’s also that ongoing regulatory scrutiny, which could change things.
The whole idea is to make it immediate and everywhere. No app install, no signup wall—just text and go.
When you send a command or pick a recipe, Poke figures out which AI model can handle it best. Then, it routes your request accordingly.
The system taps models from cloud providers and open-source projects. That way, it can handle everything from natural language tasks to automation and data retrieval.
This model-agnostic routing is at the core of Poke’s whole thing—a single chat interface that quietly orchestrates a bunch of tools and services for you.
Integrated recipes and reach across popular apps
Poke comes with a growing library of pre-made recipes that link to well-known apps and services.
- Gmail and Google Calendar for email and scheduling
- Notion for notes and knowledge management
- Strava for activity tracking
- Withings devices for health data
- Philips Hue and Sonos for smart-home control
- Developer tools and other third-party services for automation
Each recipe installs with a single click and uses standard authorization to connect Poke with your accounts. The goal? Lower the barrier to automation.
You can string together routines, like a morning briefing that summarizes meetings, workouts, and weather, all without leaving your messaging app.
Security, privacy, and the creator ecosystem
Poke puts a lot of emphasis on security with layered protections, limited permissions, and optional analytics. That’s pretty important when you’re letting an assistant access personal data and connected services.
TechCrunch pointed out that, as of publication, no independent audit had been done. That’s a fair callout—third-party reviews matter, especially in this fast-moving area.
The architecture tries to minimize data exposure and only asks for the permissions it actually needs. You can opt out of analytics if you want.
Creator-driven growth and monetization
Poke is betting big on its creator community. They encourage people to build and share automations, even offering monetary incentives for popular recipes.
This approach has helped the platform grow quickly. Thousands of user-made automations now add value to the whole ecosystem.
Business model, funding, and the roadmap
Poke uses a usage-based pricing model. Basic features that don’t require real-time processing can be free.
If you want continuous or real-time stuff—like automating every email—you’ll pay for that. The pricing is set up to encourage growth and wide adoption, not just quick profits.
They want Poke to be a utility you can’t imagine living without.
Funding and team
The startup has raised around $25 million in seed funding. On top of that, it recently closed a $10 million round, which puts the 10-person company at a post-money valuation of about $300 million.
Poke’s team is small, but they’ve got a lot of experience. They seem to have a pretty clear plan to scale, mostly by offering strong incentives for creators and plugging into more platforms.
Poke looks ready to push further into mainstream daily use for AI-driven automation. There’s a sense that they’re just getting started.
Here is the source article for this story: Poke makes AI agents as easy as sending a text