Skip to main content

Medical Information Cloud Labs

Medical Information Cloud Labs is the home for Mavens Labs capabilities built for Medical Information Cloud, optional add-on packages, installed on top of the core product, that extend MIC with emerging functionality such as AI-assisted medical information handling and integrations with the Veeva ecosystem.

What Mavens Labs is​

Mavens Labs is a program in which Mavens engineers, professional services staff, and other employees share modules and apps created through inspiration and collaboration with the client community. Labs packages:

  • are not official products but community projects, installed separately through the Install Service;
  • are not officially validated or documented as part of a Medical Information Cloud release; and
  • may receive updates outside the normal release cycle.

They are provided as is, without product-team support, and an organization that installs a Labs package takes on responsibility for evaluating it for its own use.

Why these capabilities live in Labs instead of the MIC core product​

Labs accelerators sit outside the core Medical Information Cloud product for different reasons, but the common thread is that Mavens cannot make the same guarantees about them that it makes about the validated core.

Integrations with external systems. The Veeva integrations (CRM, Network, and Vault) depend on third-party platforms that Mavens neither controls nor has standing access to. Without that access, Mavens cannot build, test, and maintain them to the service levels it holds the core product to, so they are delivered and supported as Labs accelerators.

Capabilities ahead of regulatory guidance. Agentic AI currently sits ahead of regulatory standards. The FDA has not published a standard for validating agentic AI. Rather than fold it into the validated product before such a standard exists, Mavens keeps MIC AI in Labs and tests it from first principles, measuring:

  • Accuracy, how reliably the AI extracts Adverse Events, Product Quality Complaints, and Requests from inbound interactions, scored against defined scenarios;
  • Repeatability, how stable the output is when the same input is run repeatedly; and
  • Functional behavior, that each step of the experience matches its documented expected outcome.

Keeping these capabilities in Labs protects the validated product's compliance posture. For MIC AI specifically, each implementation produces artifacts (the test plan, test fixtures, results, and a summary report covering structural and semantic pass rates, repeatability, and functional results) that a customer's compliance and validation team can use as testing evidence toward their own requirements.

note

Labs packages are optional and configurable. MIC AI, the AI add-on, runs on Salesforce Agentforce, Einstein, and Data 360, and is built around secure data retrieval, an immutable paper trail, and zero data retention. Its components use the MAI_ prefix, a naming convention, not a namespace (AIMI is unnamespaced), distinct from the core Medical Information Cloud product it extends (MED / mvn).

What's in MIC Labs​

  • MIC AI, the AI add-on (the AIMI package): the AIMI Tier 1 Agent that orchestrates agentic fulfillment, the AIMI Recommends review panel that governs it, and Ask AIMI for running agent actions on demand.
  • Veeva CRM, integration with Veeva CRM (including the signature viewer) for medical inquiry submission and healthcare professional account search.
  • Veeva Network, integration with Veeva Network for master data management.
  • Veeva Vault, integration with Veeva Vault for searching external content.