Services

One team, one use case. Then the next.

We set up Widal.AS for any team in a regulated company and run it with you in your own cloud. Every engagement follows the same path: find the first use case, prove it with one team, run it, and add the next. Builds come in when a use case needs them.

Four steps, one path.

Each step says who it is for, what we do, and what you have at the end. The end of one step is the start of the next.

1. Use case read

Who it is for
A company that wants AI agents to take on real work, and does not yet know where to start.
What we do
We talk to your teams, read their procedures and look at the systems they use. Then we pick the first team and use case: work that repeats, that a fixed check can test, and that a person can sign off.
What you have at the end
A short written plan for the pilot: the team, the use case, what the agents do, which checks test the work, and who signs.

2. Pilot

Who it is for
One team and one use case. For example: code changes for an engineering team, deviation reports for a quality team, or prior authorization requests for a patient access team.
What we do
We write the workflow with the team, set up the agents and the checks in your own cloud account, and measure the work against the number the team already uses. One of our engineers works with the team every day. For engineering teams, the pilot uses Widal.Factory. For other teams, the pilot is where we set up and test the workflow for that use case.
What you have at the end
One use case running for one team, the hub that shows its work, and a written read-out: what got done, how fast, at what quality and cost, and what to do next.

3. Run

Who it is for
A company whose first use case works, and that wants to keep it running and add the next.
What we do
We run your AI systems with you: updates, monitoring, backups and incident response. Each week we review the agents' mistakes with each team and turn them into rules and checks. When one use case works, we add the next, one team at a time. A named Widal engineer stays with you.
What you have at the end
More teams with use cases that run every day, all on one hub. A mistake found once becomes a rule or a check for every agent.

4. Builds

Who it is for
A use case that needs more than a workflow: a connection to another system, or a new agent for one job.
What we do
A small team of our engineers (a pod) builds it with the team that will use it, and owns the work from start to end. That includes connections to electronic health records (EHRs) such as Epic, Oracle Cerner, Athenahealth and AdvancedMD, and to labs and payers, in the standard formats: FHIR, HL7v2 and X12.
What you have at the end
The use case runs with the new connection or agent. Whatever we built, you keep: the code, the workflows, the checks and the documents to run it.

We give prices on a call, once we know your team and the work. Model use and cloud costs are billed by your own providers, with no markup from us.

Who does the work.

Forward deployed engineers: senior engineers who have shipped software in regulated companies. They sit with the team that uses the agent system, write its workflows and checks, and ship the use case that matters most. They are part of every engagement, and more can be added for a larger build.

Senior people, all the way through

The person on the first call is the person who builds your system.

You own what we ship

Workflows, checks, code and documents are yours. Widal.AS runs in your own cloud account.

Measured the way you measure

We pick the number you already run against, such as encounter cycle time, claims yield or alert precision, and work against it.

Forward deployed engineers

Every piece of work goes the same way.

Whatever we build for you, and whatever Widal.AS does, follows the same four steps. A person hands off work. An agent does it, inside your own cloud. Fixed checks test it. A person signs.

Tell us which work you want to hand off first.

Book a call