Want to connect with Interaction?
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.
Interaction Co. is highly relevant to the agent ecosystem because they are solving the orchestration problem. As the industry moves from single-model chat to multi-agent systems, the logic required to coordinate these agents becomes the most valuable part of the stack. Interaction Co. is building this logic and applying it to consumer-facing messaging channels.
Their commitment to building a stack that feels "genuinely useful"—as opposed to just impressive—places them at the forefront of the utility-agent movement. By focusing on WhatsApp and messaging as the primary surface, they are also exploring how agents can be integrated into high-retention habits. This makes them a key player to watch for anyone interested in the practical deployment and orchestration of multi-agent systems in the wild.
Interaction Co. is building in the specific territory where Large Language Models transition from interesting novelties to reliable tools. Their primary product, Poke, is a multi-agent orchestration stack designed to move beyond the limitations of the single-prompt chat interface. While the industry has spent the last two years focused on making models more coherent, Interaction Co. is focused on the layer above the model: the orchestration of specialized agents that can handle sub-tasks and return a cohesive result to the user.
This approach acknowledges a growing consensus in the AI community that general-purpose models often fail when asked to perform multi-step, complex workflows without structured guidance. By building an orchestration stack, Interaction Co. allows for a modular system where different agents can be tuned for specific parts of a task, whether that is data retrieval, decision-making, or user communication. This architectural choice is visible in the "OpenPoke" community implementation, which highlights a stack built for genuine utility rather than simple text generation.
One of the most telling aspects of Interaction Co.’s strategy is their use of the WhatsApp API. Most AI startups build standalone applications or web interfaces that require users to change their behavior and adopt a new destination. Interaction Co. appears to be meeting users where they already reside. By integrating with WhatsApp, they bypass the app-fatigue hurdle and position their agents inside the world's most popular messaging infrastructure. This allows for a push-based interaction model where the AI can proactively reach out to users to complete tasks or provide updates, a significant departure from the reactive nature of most web-based LLM wrappers.
Their LinkedIn activity suggests they are paying Meta for WhatsApp API usage, indicating a commitment to professional-grade distribution. This messaging-first approach is particularly relevant for agentic workflows, where a user might initiate a long-running task and expect a notification once the agents have finished their work. It shifts the AI from a synchronous chat session to an asynchronous service provider.
Interaction Co. operates with a minimalist public presence that contrasts with the typical marketing-heavy approach of Silicon Valley startups. Their primary website, interaction.co, is a simple landing page that points toward Poke and their hiring efforts. Based in California and officially known as The Interaction Company of California, they have secured premium digital real estate with the poke.com domain. This branding suggests an interest in high-frequency, simple interactions—the digital equivalent of a "poke"—that result in immediate value for the user.
By focusing on the underlying stack and distribution rather than flashy web interfaces, they are carving out a niche as an infrastructure-aware product company. They are not merely an application layer but a team building the connective tissue between advanced models and everyday consumer communication. This focus on the "interaction" itself, as their name implies, is their primary differentiator in a crowded market of general-purpose assistants.
A multi-agent orchestration stack for useful AI interactions.
Display and customize Markdown text in SwiftUI
Example MCP servers for building Poke integrations. Includes templates for no-auth, API key, and OAuth implementations.
ORM for TypeScript and JavaScript. Supports MySQL, PostgreSQL, MariaDB, SQLite, MS SQL Server, Oracle, SAP Hana, WebSQL databases. Works in NodeJS, Browser, Ionic, Cordova and Electron platforms.
Interaction is hiring
You've explored Interaction.
Join organizations building the agentic web.