MEDICALCITY · The AI Factory for Healthcare
Healthcare is our starting point. MEDICALCITY designs, develops and deploys multi-agent systems that support healthcare professionals. Health, food, housing and mobility guide the needs to explore. The platform connects knowledge, business tools and human approvals through digital infrastructure; physical execution interfaces are developed within separate robotics pilots.
For healthcare teams
Care staff · Medical secretaries · Administrative teams · Healthcare facilities · Physicians
Four pillars of life
For each need, we identify the information, tools and approvals to connect. Uses related to food, housing and mobility remain to be defined and developed with teams.
Connect reference documents, case-management tools and the professionals responsible for review, with access defined for the project.
Define meal-request sources, team instructions and the review process before developing the connections needed.
Identify reference documents for a residential admission request, follow-up tools and the people responsible for approving the information.
Define travel-related documents, people to contact and review steps to connect with the team’s tools.
Teams specify sources, constraints and approvals. Agents prepare; the people responsible check, correct and decide what happens next.
These roles complement one another around healthcare work. Features, connections and responsibilities are confirmed during scoping; physical execution belongs to separate pilots with a designated operator.
The team defines the need, usable information and the person responsible for approval. Medical decisions and signatures remain with authorised professionals. In a robotics pilot, the operator authorises and supervises each physical task.
Drafting, administrative and coordination agents prepare documents, files and follow-up. Within a pilot, they can also propose a robotic task sheet and a report for review.
Robots are considered for defined physical tasks, subject to hardware and site capabilities. The operator checks conditions and decides when to start, with stopping and takeover arrangements to validate before the trial. The proposed tasks exclude care procedures and physical contact with a person.
From digital work to physical execution
The platform connects two infrastructures developed by the AI Factory: the digital foundation for agents and the interfaces for physical robot execution. Business software remains part of the team’s toolkit; connections, permissions and execution feedback are defined for each project.
Connect specialised agents to AI models, reference knowledge, memory and business software. Workflows organise document preparation, administrative processes and follow-up, with access and approvals defined for each project.
Research direction: “Multi-Agent AGI”, towards artificial general intelligence. A development ambition; the systems presented rely on specialised agents.
Prepare robot connections, operator supervision and execution feedback. Each pilot must confirm hardware capabilities, permitted tasks, site conditions and ways to stop or take over the work.
Robotics: pilots in preparation. Drones: a longer-term development prospect. Scope, timing and service conditions remain to be confirmed for each project.
An AI draft never triggers a physical action. If information is missing or a condition is not met, the task waits for clarification and a human decision.
Collaboration example
To prepare an administrative file, agents can share the work using your information and the official list of required documents. This scenario must be adapted to the agreed scope.
Specify the file to prepare, the sources that may be used, the expected result and the professional responsible for review.
This role compares the supplied documents with the reference list and identifies what is present, missing or needs checking.
This role proposes a draft based on the supplied information, separating available details from points still to complete.
This role gathers the prepared materials, open questions and a proposed follow-up plan for your team.
You review, correct and decide what happens next. If information is missing or contradictory, seek clarification before sending anything. Drafts can contain errors; preparation does not guarantee that a file will be accepted.
Explore workflowsResponsibilities and working arrangements are defined with the people involved, based on your work and tools.
Choose the documents, instructions and checklists to use; check their date and version with the responsible professional. Define which references to keep with drafts so they can be checked more easily.
Check possible connections, the minimum necessary data, authorised access, and hosting and confidentiality requirements appropriate to the context.
Name the people responsible, the actions requiring approval and the records to keep. Specify how to correct a result, pause a task or take it over.
The AI Factory defines components, exchanges and approval points. BAYON, business software and robotic interfaces each have a role to specify and test within the project scope.
Within the MEDICALCITY architecture, BAYON has the role of a shared reference for data and rules. Authorised sources, their versions, working memory and the features used are defined for the project.
Connect models, agent skills and the applications useful to the workflow. Check each connection: required exchanges, compatibility, access, hosting and confidentiality. Available integrations are confirmed before setup.
For a pilot, define how a checked task sheet becomes a task authorised by the operator and how the result returns to the team. Test hardware limits, stopping, missing information and human takeover before the on-site trial.
Evaluate each scope with the team against usual practice: quality, errors, corrections and total preparation, review and, where relevant, supervision time. Agree on criteria to continue, adjust or stop. Features and service terms are confirmed before setup; a robotics pilot requires a separate agreement.
Start with a task related to health, food, housing or mobility. Let’s identify the sources, tools and people to connect within a scope for testing. The AI Factory scopes and develops the system; Workflows illustrates preparation and approval. Any robotics pilot has its own scope.
