Photon Declares Mobile Apps Dead, Raises $4.5M to Push AI Agents into Messaging Platforms

3–4 minutes

The App Funeral: A Symbolic Gesture in San Francisco

On September 17, Photon staged a real funeral in a San Francisco church, complete with speeches, a coffin for app icons, and panels from Vercel, Stripe, and OpenAI. CEO Daniel Tian described the event as both a tongue‑in‑cheek ceremony and a developer day, underscoring the startup’s belief that the era of traditional mobile apps is ending.

Funding Milestone and Rapid Growth

Photon announced a $4.5 million seed round co‑led by Gradient and A*, with participation from Vercel, HSG (HongShan), Z Fellows, Llama Ventures, Karman, and other angel investors. The capital arrives after the company signed up more than 40,000 developers and reported a tenfold revenue increase over four months.

The Vision: Agents Replace Apps

Photon’s thesis is that users no longer want to install new apps. Instead, they prefer agents that operate within the messaging platforms they already use, such as iMessage, WhatsApp, and others. Meta’s Muse, which has surged to the top of app store charts, is testing the same premise that users prefer agents over apps. Tian told the audience,

“Honestly, it’s going to take a while, but I think, directionally, it’s inevitable.”
He added that the shift from apps to agents represents a fundamental change in how people interact with software.

Product Offering and Developer Tools

Photon provides a managed platform built on an open‑source foundation. The open‑source version still accounts for 98 % of usage. The hosted tier guarantees 99.95 % uptime, is SOC 2 Type II and HIPAA compliant, and is sold via subscription tiers. Any project exceeding ten users must move to a paid plan. The platform includes a unified API, an extensible channel framework, a command‑line interface, and an observability suite, enabling agents to run on iMessage, WhatsApp, Telegram, SMS, RCS, email, voice, and more.

Customer and Integration Ecosystem

Photon’s platform powers agents for a range of verticals:

  • Corgi Insurance
  • Social introductions app Boardy
  • Gen Z dating platform Ditto
  • Business banking platform Rho
  • Finance AI assistant Fliptexts
  • AI email client Slashy

Integration partners include Vercel (Eve and ChatSDK), Nous Research’s Hermes agent, and Tencent’s QClaw and NanoClaw. The technology also integrates with frameworks such as LangChain, Mastra, Convex, Render, Railway, and Telnyx.

Metrics and Market Impact

While revenue figures remain confidential, Photon reports less than 3 % churn and a fivefold surge in messaging volume in the past month alone. The company estimates that its technology now reaches millions of end users across both open‑source and hosted deployments.

Future Outlook: Agent‑to‑Agent Communication

Tian envisions a next‑generation “A‑to‑A” communication layer where agents can locate and delegate tasks to other agents.

“I think, in the future, agents will be able to find you the agent that can do the right things — so it’s going to be the A‑to‑A communication layer,”
he explained. As an example, a travel‑planning agent could summon a flight‑booking or hotel‑booking agent when needed.
“It’s been pretty powerful to observe the whole A‑to‑A‑to‑P [agent‑to‑agent‑to‑person] space,”
he added.

Origins and Founders

Co‑founders Daniel Tian and Ryan Zhu, the CTO, first explored the concept while building consumer apps as students at hackathons. Tian later dropped out of UPenn’s M&T program, and Zhu left high school to become a visiting student at MIT’s Media Lab. Their early open‑source bot for iMessage attracted thousands of developers, prompting them to double down on the agent platform.

Conclusion

Photon’s dramatic funeral ceremony, combined with a $4.5 M seed round and rapid developer adoption, signals a bold bet that AI agents embedded in messaging apps will supplant traditional mobile applications. If the trend continues, the company could become the backbone for a new generation of conversational experiences, both human‑to‑agent and agent‑to‑agent.