Want to connect with Fei?
Join organizations building the agentic web. Get introductions, share updates, and shape the future of .agent.
Is this your company?
Claim this profile to update your info, add products, and connect with the community.
Fei is a direct application of agentic AI within a consumer-facing communication protocol. It represents the "surface-first" approach to agent development, where the choice of deployment environment—in this case, iMessage—is as important as the underlying model. This makes Fei a relevant case study for the agent ecosystem, particularly for developers looking to bypass the app store and reduce user friction.
Within the agent stack, Fei operates at the interaction layer. It focuses on how an agent can be integrated into existing daily habits without requiring a change in user behavior. By championing the messaging thread as a primary site for AI interaction, Group Project is pushing against the trend of monolithic AI apps. It highlights a future where agents are not destinations themselves, but rather participants in the communication channels we already use.
Fei is a mobile-first artificial intelligence assistant that operates through a single, specific interface: Apple's iMessage. While the broader market for AI agents is currently dominated by complex web applications and dedicated mobile software, Fei is a bet on the persistence of the messaging thread. It is developed by an entity known as Group Project and is marketed under the tagline "The pocket agent." This branding is literal. Because the product exists inside the user's primary messaging app, it is as accessible as a conversation with a contact.
The distribution strategy here is the most notable part of the product. Most consumer AI companies face a significant hurdle in user acquisition: the friction of downloading an app, creating an account, and learning a new interface. By utilizing the iMessage protocol, Fei removes these barriers. Users interact with the agent by sending a text message to a designated iCloud address. This allows the agent to inherit the native features of the iOS environment, including notifications, lock-screen interactions, and the general familiarity of the iMessage UI.
The company behind Fei, Group Project, maintains a minimalist presence. Their public-facing web infrastructure is almost non-existent beyond a single landing page, and many standard corporate subdirectories—such as about, team, or pricing—return 404 errors. This suggests a product that is either in an early, lean stage or one that intentionally avoids the typical venture-backed startup aesthetic. The use of an iCloud email address for the primary contact and SMS trigger points to a small, agile operation that prioritizes core agent utility over corporate scaling.
This minimalist approach is reflected in the product's identity. There are no promises of "transforming industries" or "revolutionizing workflows." Instead, the focus is on being a "pocket agent." This implies a tool for quick queries, brief tasks, and conversational assistance. It is a utility rather than a platform. In a market where competitors are building massive, multi-modal systems with complex prompt engineering interfaces, Fei is a return to the basics of chat-based interaction.
From a competitive standpoint, Fei sits in an unusual position. It does not compete directly with the large-scale enterprise agents like those from Salesforce or Microsoft. Instead, it is a competitor to the consumer-grade assistants that people use for day-to-day tasks. Its primary advantage is its lack of friction. If a user can get an answer from an iMessage thread faster than they can open the ChatGPT app and wait for it to load, the iMessage thread wins.
However, this reliance on a third-party platform like iMessage comes with trade-offs. The product is limited by the constraints of what a text-based interface can do. It cannot easily offer the rich, interactive UI components found in custom-built apps. For Group Project, this seems to be an acceptable compromise. By staying within the messaging thread, they are meeting the user where they already are, which remains one of the most difficult challenges in the consumer software market.
An AI agent accessible exclusively through iMessage.
Weather Underground Handler Service
Repository with postman requests to create swamp-fei entities.
LoRa Server configuration and deploy to work with SWAMP Project probes
Versão com serviços para dispositivos embarcados
Fei is hiring
You've explored Fei.
Join organizations building the agentic web.