We're building the runtime for agent teams.

The first wave of AI agents proved the idea: a model with tools and a loop can do real work. The harder problem showed up immediately — keeping those loops alive in production. Steps fail, models drift, budgets blow up, and a single run can fan out into dozens of calls that all need to be traced, retried, and merged.

We had built that plumbing more than once, on more than one team, and it was never the interesting part. LoopLlama is the result: describe the work, hand us a crew of agents, and trigger runs over a plain REST API. We sequence the agents, route each step to whichever model fits, persist every step's input, output, and token count, and meter usage so you only pay for the steps your agents actually run.

Since then it has grown into a full runtime for agent teams. Workflows are stateful and resumable, and every run is traceable step by step — with per-step outcome badges and diagnostics the moment something goes wrong. Human-in-the-loop gates let an agent pause for input or approval on sensitive actions and pick up exactly where it left off. Built-in connectors put real tools in an agent's hands — read and send Gmail, query and update Salesforce — governed by per-capability toggles, with OAuth tokens encrypted at rest.

Teams compose crews from an Agent Hub of reusable, prebuilt roles or their own saved templates, and schedule workflows to run on a recurring cron in their own time zone — no babysitting required. For the people operating the platform, role-based access control, an audit log, usage and cost dashboards, API-key oversight, and account suspension keep multi-tenant deployments governable.

In 2026 the runtime opened up. Any step can run on any model from any provider — Claude, GPT, Gemini, Grok, DeepSeek, Mistral, or a Perplexity Sonar step that searches the live web and cites its sources — and model usage is passed through at the provider's list price with no markup. Thirteen connectors, signed webhooks, a typed TypeScript SDK, and a durable run queue that survives deploys turned LoopLlama from an API into infrastructure teams can build on.

Today lenders, insurance agencies, property managers, recruiters, accounting firms, and legal teams use LoopLlama to put agent crews on the back-office work they already do — inside the CRM, inbox, calendar, and database they already use, with a human approving every send — without standing up an agent framework of their own.