Is Glen a data catalog or a feature store?

No. Glen is shared memory for the AI agents you use. It captures the tacit knowledge, caveats, definitions, and lessons, that catalogs and feature stores do not, and feeds it back to your agents automatically.

How do my analysis tools use Glen?

Glen is a standard MCP server. Connect any agent or assistant to it as a tool, and it reads relevant data context and records new findings in one round trip while you work.

Shared memory for data scientists

Data scientists live in a world of context that never makes it into a notebook: which table is the source of truth, why a column is unreliable before a certain date, what a metric actually means, which features blew up last time, and the dozen tacit conventions that separate a correct analysis from a misleading one. When you use AI agents to query warehouses, write SQL, or explore datasets, that context resets every session, so the agent rediscovers the same data caveats and you re-explain the same definitions. Glen, shared memory for AI agents, gives the agents data scientists use long-term shared memory as a single MCP tool, so hard-won knowledge about your data persists and is shared across the team.

Point your AI tools at Glen over MCP and any agent, whether it is writing SQL, building features, or exploring a dataset, can call one tool that retrieves relevant long-term context and records new facts in a single round trip. Before it queries, the agent pulls what the team already knows: the canonical table for revenue, the date the tracking changed, the column that double-counts, the metric definition leadership agreed on. As the work proceeds, it writes back new findings, a data-quality issue spotted, a join that finally worked, a definition clarified, so the next analysis inherits it.

The point is that this knowledge is institutional, not personal. Glen is org-scoped, so the memory is shared across every data scientist, every agent, and every tool in your organization. A caveat one analyst's agent learned is available to the whole team next time, and because Glen is a standard MCP server, the same memory is readable from Claude Code, Cursor, your warehouse-query agent, or any other MCP client. The data context that usually lives in people's heads and scattered Slack threads becomes durable, queryable memory your agents act on automatically. You connect once over OAuth or an API key and stop re-teaching the same data lore every session.

FAQ

Is Glen a data catalog or a feature store?: No. Glen is shared memory for the AI agents you use. It captures the tacit knowledge, caveats, definitions, and lessons, that catalogs and feature stores do not, and feeds it back to your agents automatically.
How do my analysis tools use Glen?: Glen is a standard MCP server. Connect any agent or assistant to it as a tool, and it reads relevant data context and records new findings in one round trip while you work.