The platform

MyLÚA plugs into women's health the way Stripe plugs into commerce.

Stripe made it so any business could take payments without building the banking machinery itself. MyLÚA does that for care. One secure system, a growing portfolio of agents, and the machinery underneath handled.

Every surface she already uses

Not another app to manage. A layer that shows up where the work happens.

A mother meets LÚA in the app, in a text, or on the web. A care team meets the same agents inside their dashboard, their site, or the tools they already run. The agent changes. The architecture doesn't.

The MyLÚA app home screen, showing the four agents a mother sees
The app: one surface, not the product

And the channels she already answers

TextEmailWeb chatSlackTeamsWhatsApp

Inside the enterprise stack, via IBM watsonx Orchestrate

The care team's surface

No new system to learn. No new tab to remember.

The same agents run as an embedded assistant inside the dashboards a care team already uses. This isn't a screenshot. It's the assistant, running here.

LÚA sits inside the system they already run. Click the lotus.

The agents

Agents are the product. The platform makes them fast and defensible.

Add the agents you need, the way you add items to a cart. Each one runs on the same secure foundation.

For mothers

  • Pregnancy Tips & SupportAnswers and support for pregnancy and postpartum.
  • Health Check-In SurveysSurveys to understand where you are emotionally, physically, and socially.
  • Find Local ResourcesSupport and programs near you.
  • Ask AnythingNot sure where to start? Share what's on your mind.

For doulas and care teams

  • Client Support & GuidanceGet help with education, planning, or client communication.
  • Client Insights & TrendsSee charts and summaries that show patterns across your clients.
  • Find Local ResourcesSupport and programs near you.
  • Ask AnythingNot sure where to start? Ask a question or describe what you need.

After every visit

The record, at the end of the visit.

A doula logs what happened in her own words, taps Generate AI Summary, and downloads a structured record of the appointment. If her agency bills Medicaid, LÚA can support notes for reimbursement (SOAP).

  • Real service time on the record: status, and the actual start to end
  • Her notes stay her words. The summary is generated from them, not instead of them

Enterprise · marketplace

MyLÚA Care Recommendation Agent

A care team submits a member's records. The agent returns a prioritized list of next steps and local resources, at the individual and population level. It's live in the IBM watsonx Orchestrate Agent Catalog, with Microsoft, Google, and AWS marketplaces to follow. Marketplaces are one way to buy. Agents can also be bought direct.

In a full deployment

She shares. The system acts. Her people stay close.

There are two ways to bring MyLÚA in. Take a single agent and run it inside the systems your teams already use, or deploy the whole app, where the agents work as one loop around her. This is that loop.

  1. She checks in. Onboarding and daily logs build a picture of her emotional, physical, and social needs over time.
  2. The system personalizes. Starter prompts, weekly activities, and notifications adapt to what she actually needs.
  3. She stays in control. Nothing reaches her doula or care team unless she has chosen to share it.
  4. Her care team acts early. What she shares becomes a next step: a resource, a check-in, a person reaching out.
A woman reading a check-in from LÚA on her phone

Security and trust

Trust is built into the architecture, not into a promise.

What she shares Logs, check-ins, appointments
Infrastructure layer PHI stops here
The AI model Receives no protected health information
Her care team Receives only what she has chosen to share

PHI never reaches the model

Protected health information never reaches the AI model, enforced at the infrastructure layer, so patient data can't leak through it.

HIPAA compliant

High-level security keeps personal information private. Data is shared only with the care team a woman chooses. It's never sold.

Consent, upfront

The consent model is opt-in from the start. The woman decides what gets shared, before anything is shared.

Patent pending

The personalization framework behind the platform is patent pending, trained under clinical data-use agreements.

Trust is why it works: 79% of users felt comfortable sharing sensitive information, and 64% completed clinical assessments.

Built with IBM watsonx

Capable agents, with the governance and scale an enterprise needs.

MyLÚA runs on IBM watsonx Orchestrate and watsonx.ai, on IBM Cloud. That pairing gives buyers capable AI agents plus the enterprise governance, security review, and scale their procurement teams expect.

Michael Conward, Ph.D., on stage at IBM THINK 2026 for a panel on enterprise AI agent governance
Watch: MyLÚA at IBM THINK 2026. On stage: enterprise AI agent governance, with IBM's VP of watsonx Orchestrate, IBM's VP of Consulting, and Snapchat's People Technology lead.

Where it goes next

Maternal care is where we proved it. Women's health is where it goes.

Each new agent is mostly configuration on a platform already live. That's how the portfolio compounds.

Live today

Maternal care

Pregnancy through postpartum, on iOS and Android, in English and Spanish. Live with Metro Mommy Agency, our first paid pilot partner, a multi-state doula network.

Next

Menopause, and beyond

A new agent, the same platform, the same PHI boundary. New surface area without a new build, which is the whole argument for having a platform.

We're live

Care is built around visits. Life happens in between.

The go-live film, in the product's own words: LÚA reaching out first, a next step instead of a score, her doula already knowing, and about 79 minutes back for every hour of care delivered.

53 seconds, sound on. The full story is in the announcement.

See the platform working.

A demo takes 30 minutes and starts from your workflow, not ours.