Gumloop is the multiplayer AI agent builder. We believe the people who know the work best should be able to automate it, and use what they build to help their whole team do more.
Anyone at a company can build, share, and optimize agents with any AI model and any integration. Teams can orchestrate those agents across their business, while IT keeps full control over data, access, and spend.
We’re a small team building as fast as humanly possible. Gusto, Ramp, Shopify, Samsara, Instacart, and Opendoor are running agents on Gumloop across their company. We’ve raised over $70M from Benchmark, First Round Capital, Y Combinator, Nexus Ventures to scale and are looking for people who care as much as we do to join the team.
Here’s what life at Gumloop is like:
•
Our beliefs – The Handbook
•
Our adventures – Wall of Photos
•
What we’re working on – Changelog
•
What our users are saying - Wall of Love
Gumloop is the AI automation platform for the enterprise. We give teams the power to easily build and deploy AI agents, in Slack, Microsoft Teams, or standalone, that automate real work across hundreds of integrations with no engineering background required. The platform has three pillars:
1.
Agents — Conversational AI assistants that use tools (integrations, workflows, code sandbox, MCP servers) to accomplish open-ended tasks. Unlike workflows that follow fixed paths, agents decide which tools to use and when, adapting based on context. Users configure agents with a system prompt, a model, and a set of tools. Agents can be deployed to a chat UI, Slack, or Microsoft Teams, and can run autonomously via scheduled or event-based triggers.
2.
Workflows — Visual, drag-and-drop automations composed of connected nodes on a canvas. Each node is a discrete operation (call an API, run AI, read a spreadsheet, send an email). Workflows are deterministic and repeatable. They can be triggered manually, on a schedule, via webhook, or by external events (new email, Slack message, etc.). Workflows can also be attached to agents as callable tools — the agent orchestrates, the workflow executes.
3.
Gumstack — Enterprise-grade MCP hosting, observability, and AI governance. Organizations provision managed MCP servers, control which AI models are available, audit every tool call, and enforce security policies across all agents and workflows.
We’re moving fast. Teams at Gusto, Ramp, Shopify, Samsara, Instacart, and Opendoor are already building and deploying AI agents on Gumloop to automate real work across their organizations. That’s not a demo. That’s production.