In a move that brilliantly blends dark tech humor with a serious bet on the future of computing, the artificial intelligence startup Photon recently hosted an actual funeral for mobile apps. Complete with a church venue, formal speeches, and a literal coffin carrying decorative app icons, the event in San Francisco on September 17 was designed to make a provocative point. Photon is so convinced that traditional mobile applications are on their way out—to be entirely replaced by conversational AI agents—that it decided to bury the app icon era in style.
The unconventional gathering was more than just performance art; it doubled as a developer day featuring panels from industry heavyweights including Vercel, Stripe, and OpenAI. More importantly, the tongue-in-cheek burial underscores a genuine commercial momentum. Photon, a company that builds infrastructure enabling developers to deploy AI agents directly over communication channels like iMessage and WhatsApp, has successfully secured $4.5 million in seed funding. The round was co-led by Gradient and A*, with additional participation from Vercel, HongShan, Z Fellows, Llama Ventures, Karman, and a roster of angel investors.
This fresh capital injection follows a period of hyper-growth for the startup. According to the company, it has signed up over 40,000 developers, experienced a tenfold surge in revenue over just four months, and seen its messaging volume multiply fivefold in the last month alone. Both its hosted and open-source iterations now reach millions of end users globally, boasting a remarkably low customer churn rate of less than 3%.
The Thesis: Moving Beyond the App Store Bottleneck
The foundational premise behind Photon stems from a frustrating realization experienced by its founders during their student days. Co-founder and CEO Daniel Tian and CTO Ryan Zhu spent years building consumer apps, often testing their mettle at hackathons. However, every time they launched a new application, they hit the exact same impenetrable wall: user acquisition and discovery. Getting people to actively visit an app store, search for an unknown utility, download it, and grant permissions proved to be an exhausting bottleneck.
Their perspective shifted when they experimented with building an agent that operated natively inside Apple’s Messages application. By leveraging iMessage, they realized they could bypass the app store entirely, reaching users directly through a utility they already opened and used dozens of times a day.
Tian initially built a rudimentary bot designed to text his friends back as him over iMessage. When he and Zhu published the underlying technology as open source on GitHub, it went viral, catching the attention of thousands of developers who immediately adopted it. The overwhelming response prompted life-altering decisions: Tian dropped out of the Jerome Fisher Program in Management and Technology at the University of Pennsylvania, while Zhu left high school early, later serving as a visiting student at MIT’s Media Lab.
Today, while Photon has evolved into a commercial platform, the open-source version of its software still accounts for a staggering 98% of its total usage.
Bridging Developers to Consumer Messaging Platforms
The core challenge for developers looking to capitalize on the conversational AI boom is fragmentation. While consumer-facing experiments like Meta’s Muse have climbed to the top of consumer app store charts by testing similar messaging-first theories, an entire ecosystem of independent developers wants to embed their applications into existing communication channels without building custom interfaces from scratch.
This is the exact market Photon targets. The startup provides a comprehensive suite of developer tools, including a unified API, an extensible channel framework, a command-line interface, and robust observability suites. Together, these tools allow developers to build and manage agent experiences that function smoothly across iMessage, WhatsApp, Telegram, SMS, RCS, email, and voice channels.
In April, the founders launched a fully managed, hosted version of the platform offered through tiered subscriptions. Unlike running the open-source code locally on a personal Mac for a hobby project, the enterprise-grade hosted platform guarantees 99.95% uptime. Furthermore, it is SOC 2 Type II and HIPAA-compliant, a crucial compliance milestone that unlocks strict enterprise and healthcare agent use cases where data privacy and security are paramount.
Photon’s utility has already attracted a diverse roster of clients spanning fintech, consumer AI, insurance, and corporate productivity. Companies utilizing the infrastructure include Corgi Insurance, social introductions app Boardy, Gen Z dating platform Ditto, business banking platform Rho, financial AI assistant Fliptexts, and AI email client Slashy, among others.
Beyond direct customers, Photon has woven itself deeply into the broader AI infrastructure web. The platform maintains key technology integrations with Vercel—which utilizes Photon components for tools like Eve and ChatSDK—and Nous Research, whose Hermes agent relies on Photon as its default iMessage integration layer. Photon also serves as the underlying layer for Tencent’s QClaw and NanoClaw projects, alongside infrastructural integrations with LangChain, Mastra, Convex, Render, Railway, and Telnyx.
Looking Ahead: The Agent-to-Agent Economy
While the current landscape is heavily defined by AI agents communicating directly with human users, Daniel Tian believes the industry is already moving toward a much larger architectural shift. He envisions a future where humans do not interact with dozens of different agents individually; instead, a primary personal assistant agent will dynamically discover and coordinate with other specialized agents behind the scenes.
Describing this emerging paradigm as an agent-to-agent communication layer, Tian points to everyday scenarios like planning a trip. Rather than navigating separate tools, a user could chat with their primary AI assistant, which would autonomously reach out to and coordinate with specialized flight-booking agents, hotel-reservation agents, and local itinerary planners. Observing the evolution of this agent-to-agent-to-person ecosystem, Tian remains convinced that while the transition away from traditional apps will take time, the directional shift is completely inevitable.
Leave a Reply