Photon Secures $4.5 Million to Build Agent Infrastructure for Messaging Apps

Photon, an artificial intelligence startup that creates agent tools for messaging platforms like iMessage and WhatsApp, has raised $4.5 million in seed funding. The investment follows a dramatic marketing stunt earlier this month, where the company held a funeral for mobile apps in a San Francisco church, symbolizing its conviction that AI agents will eventually supersede traditional applications.

Co-founder and CEO Daniel Tian described the September 17 event as both a symbolic farewell to the app era and a functional developer day featuring panels from industry players such as Vercel, Stripe, and OpenAI. Despite the humorous tone, Tian maintains that the transition from apps to agents is inevitable, noting that while the shift will take time, the direction is clear.

The fundraising round was co-led by Gradient and A*, with participation from Vercel, HongShan, Z Fellows, Llama Ventures, and Karman. The capital will support the development of Photon’s suite of tools, which includes a unified API, a command-line interface, an observability platform, and a channel framework. These resources enable developers to create agent experiences that operate across multiple channels, including Telegram, SMS, RCS, email, and voice.

Photon’s origins trace back to the experiences of Tian and CTO Ryan Zhu, who built consumer apps as students. They frequently encountered the same obstacle: user discovery. They realized that building agents within existing messaging ecosystems, particularly Apple’s iMessage, could bypass this bottleneck by reaching users through apps they already used daily. Their open-source project gained traction on GitHub, leading Tian to drop out of the University of Pennsylvania’s M&T program and Zhu to leave high school for a visiting position at MIT’s Media Lab.

Currently, the open-source version of Photon accounts for 98% of its usage. In April, the founders launched a managed platform offering higher reliability, including 99.95% uptime and SOC 2 Type II and HIPAA compliance, which opens doors for healthcare applications. Since then, the company has accumulated over 40,000 developer sign-ups and reported a tenfold increase in revenue over four months, with churn rates below 3%. Messaging volume has also quintupled in the last month.

Notable customers include Corgi Insurance, social app Boardy, dating platform Ditto, banking platform Rho, and AI assistant Fliptexts. Technological integrations include Vercel’s Eve and ChatSDK, Nous Research’s Hermes agent, and Tencent’s QClaw and NanoClaw. Framework connections span LangChain, Mastra, Convex, Render, Railway, and Telnyx.

Looking ahead, Tian envisions a future where agents communicate directly with one another, creating an A-to-A communication layer. He anticipates scenarios where a personal AI assistant could autonomously interact with flight-booking or hotel-booking agents to facilitate complex tasks, marking the evolution into an A-to-A-to-P (agent-to-agent-to-person) model.