Maya · Operations 9:14 AM
Prepare tomorrow's leadership brief. Compare the pipeline report with last week's decisions, then flag anything without an owner.
Customer-owned AI work systems
Foundry 41 installs and operates an AI work system in your cloud account. A request in messaging starts an isolated job with the company context and tools you approve, then returns files, decisions, or code for a person to review. You keep the system.
The coordination problem
Giving everyone a chatbot can speed up isolated tasks while scattering context, decisions, and outputs across private sessions. Foundry gives the company one governed way to delegate and review real work.
Operations, finance, and owners work from separate chats, document copies, and business systems. Context splinters and decisions disappear.
One request, shared approved context, a named reviewer, and a visible outcome connected to the systems the company chooses.
What we install
We currently call the installed layer Foundation. It combines company context, bounded tools, isolated execution, human review, and the Software Factory that lets the system improve safely.
01 · COMPANY BRAIN
The Company Brain keeps skills, memory, configuration, and approved sources in a form your team can inspect. Each job receives only the relevant context and access.
Approved policies, team memory, system data, and a versioned workflow skill.
The task receives only the relevant context and access—not the whole company brain.
An isolated job completes the task and returns an output for a person to review.
02 · ISOLATED EXECUTION
Each conversation gets a dedicated job environment. The agent receives bounded capabilities, performs the work, and reports back with an artifact or decision—not an invisible chain of autonomous actions.
03 · HUMAN CONTROL
Rules decide what the agent may prepare. Named owners handle exceptions, approve consequential actions, and keep the reasoning with the work record.
The file is complete except for an authorization date outside the approved rule. The draft is saved; the practice system has not been changed.
A concrete first workflow
This illustrative clinic example shows the end result: not “chat with AI,” but a visible case that moves from intake to preparation, human judgment, and system outcome.
REFERRAL FOLLOW-UP
The agent prepares routine work. Staff see exceptions and approve action. The record keeps the context, draft, decision, and outcome together.
Approved referral fields and source records enter one work record.
Packet checked, missing information flagged, drafts saved.
A named owner reviews the one item outside the approved rule.
Only after approval; the work record retains the decision and output.
Why customer-owned matters
The installed system runs in your cloud account, under your keys. You pay model, cloud, API, and software providers directly. Foundry works through delegated access you can revoke, then operates the system or hands it over.
Talk through the architectureThe deployed system, operating skills, connectors, and release history.
Data, memory, configuration, and the rules that govern access.
Your provider relationships and secrets remain under your control.
Decisions, logs, files, code changes, and outputs stay with the company.
Revoke Foundry's role or use the runbook to operate the system yourself.
How the engagement works
For a scoped implementation with access and integration readiness, each stage leaves a usable artifact in your environment. The 30-day window is a target, not a black-box promise.
Days 1–5
Map the problem, systems, permissions, owner, approval points, and definition of production-ready.
Days 6–20
Stand up your environment, approved integrations, Company Brain context, and the first workflow.
Days 21–30
Evaluate, test approvals, release with monitoring and rollback, and leave an operating backlog.
After launch
Foundry monitors and improves the workflow, or your team takes the runbook and operates it.
SCOPED IMPLEMENTATION
Install the owned system and put one important end-to-end workflow into production.
ONGOING OPERATION
Keep the system healthy and improve the next highest-value part of the workflow.
Built from working software
Sky is Foundry 41's internal reference implementation. A messaging thread starts a job in a dedicated sandbox with controlled access to code and capabilities. It selects approved context, performs the work, and reports back with reviewable outputs.
Selected outputs—memory, code changes, files, todos, and routines—can survive a session. Sky and Grow, a second internal instance, are not customer deployments. Broader work identity and automatic resume remain architectural direction, not a production claim.
Bring one operational problem
Show us a workflow that crosses people and systems. We will map the context, approvals, integrations, and production path—and tell you whether it fits a 30-business-day implementation.