Home / Responsible AI Planning for a Digital Health App MVP with an AI Discovery Workshop

Responsible AI Planning for a Digital Health App MVP: AI Discovery Workshop

Aiming for AI in an App? Starting with an AI Discovery Workshop

Asterisk Health is building a platform to support better outcomes in cardiovascular care. With deep knowledge in healthcare and diagnostics, their founders understood the potential of AI, but didn’t yet know how to apply it to real product features. Their challenge was clarity: What could AI actually do for their users and their product?

To avoid wasting resources or making risky claims, they wanted expert input early in their journey. Their goal was to get clarity on the actual feasibility of AI features and define which ideas were worth building and which ones weren’t. That’s when they reached out to us.

Challenge

Unclear where AI could bring value to the product

No internal technical capacity to assess feasibility

Uncertainty on regulatory constraints and MDR

Solution

A structured AI Workshop individually tailored

Identified 8 AI use cases and prioritised top 3

Mapped integration options for short- and mid-term

Outcome

Gained clarity on realistic AI opportunities

Defined a strategic direction for product innovation

Prepared internal team for the next development phase

How to Explore AI Without an In-House Tech Team

Asterisk had no internal AI engineers or data scientists, but they wanted to explore AI’s potential to improve their digital health product. What they needed was a clear, realistic evaluation of whether AI could provide tangible benefits and how to scope it properly. And that before wasting money on development or overpromising results.

They came to us to guide them through this process. We proposed a focused AI Discovery Workshop for their app idea, specifically designed for non-technical teams to help them identify, assess and prioritise AI opportunities. This took into account their specific data availability, user value, feasibility, and MDR constraints.

A Structured AI Discovery Workshop to Define Value

We kicked off the collaboration with a tailored workshop designed to help Asterisk explore the potential of AI in their digital health product. The session was structured around the specific challenges and goals they shared with us in advance. To guide the discussion, we prepared a set of relevant AI use cases tailored to their needs.

First, we analysed the current pain points and outlined some realistic use cases for AI. Next, we discussed the type and quality of data required. The most promising ideas were prioritised using clear evaluation criteria: technical feasibility, added user value anddata readiness. This helped the team understand which options could be implemented immediately, which would require further work and which should be discarded.

"A whiteboard titled “AI Use Case Canvas” displaying categorized AI healthcare applications such as “Symptom Checker,” “Intake Automation,” “Lifestyle Nudges,” and “Triage Aid.” Sticky notes mark mapped features across different quadrants. The image reflects a collaborative session from an AI Discovery Workshop aimed at defining viable and compliant AI use cases in digital health."

Compliance, Safety and Risk for AI in Healthcare

Understanding the regulatory implications of using AI in a digital health app was a key part of the AI Discovery Workshop. We explained how the MDR, GDPR and AI Act apply to medical software and discussed what this means for development planning and documentation.

We outlined the requirements for AI classification under MDR, including safety risk assessment and traceability. By identifying early what kind of data, validation and documentation would be required, Asterisk could better understand the long-term compliance implications of their ideas.

Planning an AI-Enabled Product Strategy

The final part of the AI Discovery Workshop focused on turning validated ideas into an actionable roadmap. We outlined the necessary development stages, from feasibility and PoC to prototyping and testing and estimated what would be needed in terms of data, tech stack and team roles.

We also identified areas where Asterisk’s in-house capabilities would require external support and foundations (like data pipelines or backend systems) that should be established early on to enable scalable AI integration. Asterisk's roadmap was complete and their AI idea ready to be developed.

What does the Client has to say?

"We came to Punktum with a lot of open questions on how to approach AI safely in women’s health, as well as how to shape Astra's 'mind' so she delivers value to our users while aligning with regulatory guardrails.

Their AI Discovery Workshop helped us design how our App "Astra" should behave, what risks to manage, and how to structure safe clinician summaries.

The workshop aligned UX with our Development team, closing a critical gap in understanding and communication. I’d recommend the workshop to any team building an AI product in a sensitive domain."

What Comes Next: From Ideas to Execution

Following the AI Discovery Workshop, the client gained clarity on how AI could be implemented in their app and what was realistically feasible with their current data and infrastructure. Their next steps include setting up technical infrastructure, validating data availability and aligning cross-functional teams around the selected use case.

With a clear understanding of feasibility, regulatory requirements and implementation complexity, they can now invest confidently into AI development without wasting time on unclear or unrealistic ideas. Punktum remains available as a technical partner should they choose to move forward.