A new startup wants to make AI agents as simple as sending a text. Poke, built by The Interaction Company of California, launched publicly in March as a personal AI assistant you can access through iMessage, SMS, Telegram, and in some markets, WhatsApp, according to TechCrunch AI.
The pitch is straightforward: instead of installing software, managing dependencies, or wrestling with terminal commands, you just text Poke what you need done. No app download required. Visit Poke.com, enter your phone number, and you’re in.
What Poke Actually Does
Poke positions itself not as another chatbot you ask questions, but as an agent that takes action. Here’s what it handles right now:
- Daily planning and calendar management – schedule meetings, get morning briefings
- Health and fitness tracking – integrates with Strava, Oura, Fitbit, Withings
- Smart home control – works with Philips Hue, Sonos, and others
- Email monitoring – alert you to specific messages from specific people
- Custom automations – write your own in plain text and share them with friends
- Photo editing, medication reminders, weather alerts, sports scores – the everyday stuff
The system uses pre-made “recipes” spanning health, productivity, finance, scheduling, travel, and developer tools. Each recipe connects to services like Gmail, Google Calendar, Outlook, Notion, Linear, Granola, and others. Developers get integrations with PostHog, Webflow, Supabase, Vercel, Sentry, GitHub, and Cursor Cloud Agents.
The OpenClaw Comparison
TechCrunch AI frames Poke as potentially “an OpenClaw for the rest of us.” That’s a meaningful comparison. OpenClaw-style agentic systems are powerful but demand technical chops. Installing through terminal, managing dependencies, troubleshooting errors, dealing with deep system access and the security concerns that come with it. For most people, that’s a non-starter.
Poke co-founder Marvin von Hagen told TechCrunch that the product evolved from watching beta testers push an earlier email-only assistant way beyond its intended scope. “People wanted to use Poke for everything,” he explained. “They started asking Poke to remind them to take their medication. They asked Poke about sports results.” The team pivoted to make Poke general-purpose, more proactive, and more personable.
Model-Agnostic by Design
Under the hood, Poke routes each task to whichever AI model fits best, whether from major providers or open source. Von Hagen sees this as a long-term advantage: “Almost all of our competitors are just big tech and labs that are bound to a specific provider. Like Meta AI will only ever be able to use Meta models, and ChatGPT will only ever be able to use OpenAI models.”
For messaging platform integration, Poke uses Linq, a solution that lets AI assistants live inside messaging apps natively.
Funding and Valuation
Poke is well-capitalized for a 10-person team. The company recently added $10 million on top of last year’s $15 million seed round, backed by Spark Capital, General Catalyst, and angel investors. Post-money valuation: $300 million.
The WhatsApp Problem
One notable limitation: WhatsApp access is restricted. Meta barred third-party general-purpose chatbots last fall. Regulators in the EU, Italy, and Brazil have opened antitrust probes over this decision, which has already brought Poke back to Brazil. Von Hagen describes Meta’s high fees for access as “malicious compliance” and expects the situation to improve.
What’s significant here isn’t just another AI assistant. It’s the distribution play. Meeting users where they already are, inside their messaging apps, removes the biggest barrier to AI agent adoption: friction. Whether Poke can deliver on the promise of a truly useful text-based agent at scale remains to be seen, but the approach is sound.
More details are available in the full report from TechCrunch AI.