Client Memory System — per-client living context, auto-loaded by channel
Walk into any client conversation already knowing everything the marketing team knows.
What was getting in the way.
Client context lived in people's heads and scattered threads. Every question started with archaeology: what did we promise, what's their setup, what happened on the last call? New team members (and the AI itself, which wakes fresh each session) had no fast path to "everything we know about this client."
Context
The full active roster of local-service-business clients.
How the work runs.
- 01
Trigger
continuous — any activity in a client channel, plus ongoing maintenance.
- 02
channel-to-client mapping links each client's
A channel-to-client mapping links each client's chat channel to their memory file; entering a client channel auto-loads that client's context.
- 03
Each file tracks the client's
Each file tracks the client's key facts, service setup, email-thread summaries, decisions, and active issues.
- 04
As work happens (calls ingested
As work happens (calls ingested, tickets closed, emails handled), the agent appends to the client file — records stay current without anyone "doing documentation."
- 05
Other workflows (cancellation monitor, portfolio
Other workflows (cancellation monitor, portfolio capture, kickoff docs) read these files as their context source.
- 06
Outcome
a self-maintaining institutional memory, one file per client.
Evidence from the workflow.
Each system has a role.
Intake / signal
Team chat (client channels)
Record of truth
Client memory files
Intake
Email (client group addresses)
Why this is Operator.
Owns a workflow end to end, running the process inside the authority you set.
Client files are internal-only and never quoted verbatim to clients. Memory is loaded only in appropriate contexts (client-specific channels), not in shared/group settings where it could leak.
Effects / Outcomes
15 active clients currently mapped channel→memory-file, each with a group email and standing context record.
Context-loading time for any client question dropped to effectively zero — the file is read before the first reply.
Downstream workflows (retention, portfolio, onboarding) run off the same records, so context compounds instead of fragmenting.
Find where a workflow like this fits.
Start with the systems, work, constraints, and authority already present in your operation.