All case studies / Production
Lunalta.ai: an AI health companion for menopause.
A doctor-validated AI health companion for women going through menopause, with medical guardrails and a clinician portal, built and run by SIEL.
- Client
- Lunalta.ai
- Sector
- Healthcare, women's health product
Growth goal
Give women going through menopause reliable answers between appointments, and give clinicians their repeated questions back.
Operating constraint
Guidance came in short appointments and long waits, and clinicians answered the same questions all day.
Measured result
A live product with a doctor-validated knowledge base and every consequential answer traceable to its source.
Lunalta had to be safe before it was clever. SIEL built the guardrails first: every answer traceable to content our doctors validated, and a clear handover to a person when it matters. They still run it with us.
The full story
- Baseline
- Women waiting for answers about symptoms and treatment they needed now. Clinicians repeating the same guidance.
- System
- A retrieval-grounded assistant over clinician-validated content on menopause, strict scope limits, escalation to a person, and a portal where clinicians oversee patients and validate content.
- Controls
- Answers stay inside validated content. Anything outside scope or consequential escalates to a person. Every answer traces to its source.
- Result
- A live product built and run by SIEL, with a doctor-validated knowledge base and traceable answers.
In depth
Women needed answers between appointments
Lunalta.ai set out to solve a familiar, high stakes problem. Women going through menopause had symptoms, worries and treatment questions at night, at weekends and in the long gaps between clinical appointments. They turned to search engines, forums and group chats, then tried to decide which advice to trust while dealing with hot flushes, sleep problems, mood changes or new medication. Clinicians saw the other side. They spent large parts of each day repeating the same safe, standard guidance and correcting whatever the internet had suggested the night before.
Lunalta.ai wanted to give women reliable, contextual answers when their clinician was not in the room and to give clinicians time back from repeated questions. They were building a product that serves both women and clinicians in a healthcare setting, so the operating standard had to be clinical, not consumer. Any system had to respect clinical judgement, keep guidance inside validated content and give clinicians confidence that they could stand behind what it said. That need for safety and control was the limit on how far they could grow.
The existing workflows could not safely carry more demand
The baseline was fragmented and slow. Guidance lived in clinical guidelines, notes, written leaflets and the experience in each clinician’s head. Women met that experience in short appointments and occasional follow up calls. Outside those moments, they were on their own, waiting and worrying. Lunalta.ai could see clear demand for something always on. Their team could not manually write, update and maintain every answer a woman might need across symptoms, treatments and personal situations. Any attempt to scale with static content alone would add review work for clinicians without taking repeated questions off their plate.
Clinicians also needed visibility and control. They could not accept a black box that generated menopause advice with no clear source. They needed to know where each answer came from, what scope it stayed within and when they would be pulled in. Without that, an AI companion would create more risk and more manual checking, not less. The opportunity was obvious. The existing tools and workflows could not stretch to meet it without a connected system that put clinical standards at the centre.
SIEL built a guarded assistant and a clinician portal
SIEL worked with Lunalta.ai to design the operation before the model. The core of the system is a retrieval grounded assistant that only answers from clinician validated content on menopause. Lunalta.ai’s medical team curate and approve the material that sits in the knowledge base. When a woman asks a question, the assistant retrieves relevant pieces of that content and composes an answer that stays inside what has been validated. If the question falls outside that scope, or touches on something that should not be automated, the system does not guess or generalise. It escalates.
Around that core, SIEL built the workflow and controls that make the companion usable in a clinical context. There are strict scope limits so the assistant knows when not to answer. Anything identified as consequential or outside scope routes to a person instead of producing a free text response. On top of this, SIEL delivered a portal where clinicians can oversee interactions, review how the assistant is used, and validate or update content in the knowledge base. The pieces fit together into one operation that serves women directly while keeping clinicians close to care and content.
Guardrails keep clinicians in control of care and content
Control is built into the way the system runs. Answers stay inside the clinician validated menopause knowledge base, so clinicians know the companion is not improvising medical guidance. Every consequential answer is traceable back to its exact source material. That traceability works in both directions. From an interaction, a clinician can see which validated documents were used to compose the answer. From a piece of content, they can see where it has been referenced. If needed, they can adjust or withdraw materials in the portal, and those changes flow through to future answers without rewriting responses one by one.
Escalation happens automatically when a question falls outside the defined scope or is marked as consequential. Instead of attempting a confident answer, the assistant directs the woman to appropriate next steps or flags the case for a human to pick up, in line with local practice. Clinicians have a clear view of how often this happens and what kinds of questions trigger it. Ownership of clinical standards and content stays with Lunalta.ai’s medical team. SIEL’s work is the connected system around them that applies those standards consistently, for every user and every question.
A live companion and a pattern for similar healthcare products
The result is a live product built and run by SIEL, used as an AI health companion for menopause. Women get consistent, doctor validated answers between appointments instead of searching in the dark or waiting for the next slot. Clinicians see fewer repeated routine questions in their day, and when the system does answer, they can see exactly which validated content it drew from. Every consequential answer is traceable to its source. That traceability gives the clinical team confidence to keep improving the knowledge base over time, knowing how changes will flow into what women see.
For a healthcare company in a similar position, the lesson is that you do not have to choose between reach and control. With a retrieval grounded assistant over clinician validated content, strict scope limits, clear escalation rules and a portal for clinical oversight, you can extend reliable guidance to more patients without stretching appointments or adding unstructured risk. SIEL’s role is to build and operate that connected system so your clinicians stay in charge of standards, while your product serves more people with the consistency they expect.
Fixed fee, agreed before we start. A senior engineer replies within two business days.
