Designing an AI-powered Recommendation System that turns health reports into personalized action

Diagnostics • consumer app

Powered by AI
Summarize this article

A unified upload experience for lab reports, prescriptions, and scans

00

Context

00

After successfully helping users understand their lab reports through AI-generated health insights, we noticed something interesting.

Users appreciated knowing what was wrong, but they still struggled with the most important question:

"What should I do now?"

Reading insights doesn't improve health.
Taking action does.

This realization led us to design the second phase of Health Insights Hub—a recommendation system capable of converting medical insights into personalized health plans.

Unlike traditional recommendation engines built using rigid medical rules, our goal was to build a scalable AI-assisted recommendation framework that could continuously evolve with new medical knowledge, user preferences, and healthcare services.

01

The Problem

01

Initially, recommendations were generated through a rule engine.

If LDL was high,
show Omega-3.

If Vitamin D was low,
show Vitamin D supplements.

If HbA1c increased,
recommend exercise.

While medically correct, the experience had several limitations.

Recommendations looked generic.

Every user received almost identical advice.

There was no explanation of why a recommendation appeared.

Adding new recommendation pathways required engineering work and medical validation every single time.

Most importantly, recommendations felt disconnected from the user's personal health story.

02

Approach

02

We interviewed 35 users who had recently viewed their AI-generated health insights.

The objective was to understand how people perceived recommendations after understanding their reports.

The findings completely changed our direction.

Users didn't trust AI with medical decisions

Almost half of users said they would never purchase supplements solely because AI recommended them.

Supplements were viewed as medical interventions rather than informational suggestions.

This meant recommendations required a stronger trust layer.

Doctor validation became equally important as AI intelligence.

Diet and lifestyle recommendations were highly appreciated

Most users felt comfortable following dietary and lifestyle advice because they were perceived as safe, actionable, and low-risk.

This suggested that behavior change should become the first recommendation layer before supplements or medications.


Recommendations felt generic

Many participants believed recommendations were generated by AI without understanding their specific condition.

Users repeatedly asked:

"Why am I seeing this recommendation?"

We realized personalization wasn't just about generating different advice.

It was about making users understand why that advice belonged to them.

Users wanted recommendations connected to their worst health issue

Long paragraphs explaining nutrition or supplements were rarely read.

Users preferred:

  • visual guidance

  • structured plans

  • simple checklists

  • clear priorities

Research showed recommendations needed better information architecture, not more information.


03

Defining the Design Challenge

03

We weren't simply designing another recommendation page.

We were designing a recommendation platform that needed to answer three questions simultaneously.

Medical Question

What is scientifically appropriate?

Personalization Question

Why is this recommendation relevant to this user?

Product Question

How do we make recommendations reusable across the entire product instead of existing on one page?

04

Existing Approach

04

The conventional healthcare approach looked like this.

Medical team creates protocols

Engineering builds rule engine

Design creates screens

Recommendation appears

Every new recommendation category required rebuilding logic.

This process took months.

It also limited experimentation.

04

Reframing the Problem

04

nstead of asking

"How do we generate recommendations?"

we asked

"What are recommendations fundamentally made of?"

That question changed the project.

04

Breaking Recommendations into Building Blocks

04

We analyzed hundreds of recommendation examples across medicine, nutrition, preventive care, and consumer health.

Instead of treating recommendations as paragraphs of text, we decomposed them into reusable UI patterns.

Examples included

  • Supplement recommendation

  • Medication warning

  • Food replacement

  • Foods to add

  • Meal plate composition

  • Lifestyle checklist

  • Exercise routine

  • Sleep protocol

  • Due test reminder

  • Advanced diagnostic recommendation

  • Doctor consultation

  • Monitoring timeline

  • Follow-up reminders

Each pattern became a reusable design component.

We called these our recommendation "lego blocks."


04

Teaching AI the Design System

04

Rather than asking AI to invent interfaces, we taught it how each component worked.

Every block contained metadata describing

  • purpose

  • interaction

  • medical intent

  • constraints

  • supported content

  • hierarchy

  • when it should appear

  • when it should never appear

Instead of generating UI,

AI selected and assembled existing product components.

This dramatically improved consistency while keeping recommendations personalized.

04

Expanding Medical Possibilities

04

Research showed users had different mental models.

Some wanted

"Tell me the single most important thing."

Others wanted

"Give me a complete improvement plan."

Others preferred

"Only show me diet."

Instead of forcing one journey, we designed multiple recommendation pathways from the same recommendation engine.

Users could explore recommendations by

  • overall priority

  • supplements

  • diet

  • lifestyle

  • tests

  • doctor consultation

The recommendation engine adapted presentation without changing underlying medical logic.

04

Designing for Trust

04

The biggest product challenge wasn't personalization.

It was trust.

Research consistently showed users trusted doctors significantly more than AI for higher-risk decisions.

We redesigned recommendation hierarchy around trust.

Lower-risk interventions like diet and lifestyle appeared first.

Supplements were paired with doctor consultation when appropriate.

High-risk biomarkers surfaced stronger clinical recommendations.

Recommendations also explained why they appeared, connecting every suggestion back to specific biomarkers and report findings.

This shifted recommendations from "AI advice" toward "health guidance."

04

Outcome

04

The project fundamentally changed how recommendations were generated inside Health Insights Hub.

Instead of relying solely on hardcoded medical rules, we created an AI-assisted recommendation framework that combined structured medical knowledge with reusable product components.

The same recommendation engine could now power

  • Full recommendation pages

  • Biomarker pages

  • Insight cards

  • Future follow-up journeys

  • Historical health tracking

  • Doctor consultation experiences

Most importantly, the architecture became extensible.

Adding new recommendation categories no longer required redesigning the entire product.

04

Key Design Contributions

04

Led end-to-end product design for AI-powered recommendations.

  • Conducted synthesis of qualitative user research to identify trust and personalization gaps.

  • Defined a modular recommendation design system reusable across multiple health surfaces.

  • Partnered with AI engineers to establish a component-driven prompting strategy instead of free-form generation.

  • Designed contextual recommendation experiences embedded directly within biomarker detail pages.

  • Created scalable interaction patterns that balanced AI flexibility with medical safety.

Create a free website with Framer, the website builder loved by startups, designers and agencies.