Shared memory for Zapier users
Zapier connects thousands of apps with triggers and actions, and its AI steps let you add language-model reasoning inside a Zap. But a Zap is stateless by design: each run fires, does its work, and forgets. The AI steps in your Zaps have no memory of prior runs, no shared knowledge about a contact or a deal, no durable record of what your automations have already decided. To work around that you wire in Storage steps, lookup tables, or an external database, and every Zap that needs memory rebuilds the same scaffolding. Glen, shared memory for AI agents, gives your Zaps one durable, org-shared memory they all read from and write to through a single MCP tool.
Hook Glen up as an MCP tool and any Zap with an AI step can retrieve relevant context and record new facts in one call. A Zap that triages an incoming email can pull what your organization already knows about the sender before the AI drafts a response; a later step can write the outcome back so the next Zap, in another workflow entirely, starts with that context already loaded. You stop juggling Storage steps and external lookup tables just to give your automations a memory, and you stop letting each Zap reinvent its own private state.
Because Glen is org-scoped, the memory is shared across every Zap you run and across the rest of your stack. A lead-routing Zap learns a durable fact about an account; a follow-up Zap reads it next time without you copying data between steps. And because Glen is a standard MCP server, the memory your Zaps write is the same memory available in Claude, Cursor, or any other MCP client, so your unattended automations and your hands-on work feed one shared knowledge base instead of drifting apart. You connect it once with OAuth or an API key and let the memory compound across thousands of runs. For people who automate their whole company in Zapier, Glen turns stateless Zaps into ones that actually remember.
FAQ
- How does Glen fit into a Zap?
- Glen is a remote MCP server. Connect it as an MCP tool and call it from an AI step to retrieve context and write observations, replacing ad-hoc Storage or database steps.
- Is the memory shared across all my Zaps?
- Yes. Glen is org-scoped, so every Zap, and every other MCP client your team uses, reads from and writes to the same memory store.