Spring Health

2025 – 2026

AI Strategy · Member Experience

From
features
to care

TL;DR

Led the strategic shift from isolated AI features to a unified, continuous care experience.

Identified fragmentation across the member journey as AI features scaled
Ran early experiments that increased bookings via AI-guided assessment entry points
Synthesized learnings across multiple pilots: intake, post-session, in-between
Defined a 0→1 product vision now shaping Spring Health's AI roadmap

My role

0→1 Strategy & Vision Lead, from early experiments through vision definition and ongoing 2026 expansion

Partners

Product, Engineering, Clinical, Content, Research, Marketing, CxO

Outcome

24K+ appointments driven. Vision now shaping Spring Health's AI roadmap and product strategy.

The problem

AI was emerging quickly, but fragmented.

AI at Spring Health was growing fast, but without coordination. Point solutions for provider match, intake, session summaries, and member support each existed in isolation. There was no shared memory, no context carried forward, and no continuity across the journey.

Members experienced AI as a collection of disconnected features, not meaningful support. This limited both member outcomes and business impact: missed conversion moments, inconsistent tone, and engagement that didn't compound over time.

Point solutions (provider match, intake, summaries) existed in isolation
No shared memory, context, or continuity across the member journey
Members experienced AI as disconnected features, not meaningful support
Lack of continuity reduced trust and long-term engagement
Fragmentation meant conversion and retention opportunities were consistently lost

The opportunity

What if AI wasn't a feature, but a relationship?

Instead of optimizing individual features, I saw an opportunity to unify AI into a single, continuous care layer spanning the entire member journey, from onboarding through ongoing support. This meant shifting the frame entirely.

From

Episodic interactions

Each AI surface solves its own moment, then disappears. Members start over every time they need help.

Ongoing relationships

AI carries context forward, knowing where a member has been and meeting them where they are now.

And

System-level design

Moving from feature-based thinking to a shared orchestration layer across the entire experience.

Phase 1: Early signal

Validating that AI can drive member bookings.

We introduced AI at the highest-intent moment in the member journey: immediately after assessment results. The hypothesis was direct: if we guide members at this moment, we can reduce drop-off and increase care bookings.

Members could describe what they were looking for in their own words and receive personalized care recommendations immediately. The experience was designed to feel like talking to someone who understands, not filling out another form.

Flow sketch: assessment results through AI companion to scheduling

Flow sketch: mapping the path from assessment results, through the AI companion, to provider selection and booking

Three paths considered, mapping Care Guide entry points against funnel performance and member experience tradeoffs

Storyboard exploration of how Care Guide surfaces within the assessment results experience

This became the first proof point that AI wasn't just assistive. It could drive core business outcomes for members and the company.

Phase 2: Expanding experiments

Good signals. Fragmented picture.

Following the early signal, we partnered across teams to explore AI in multiple moments across the member journey: provider fit, guided intake, session summaries and takeaways, interactive journaling, and in-the-moment support.

Each experiment performed well individually. But looking across all of them, a different pattern emerged: each feature was creating its own silo, with duplicated logic, inconsistent tone, and no memory between them. We weren't building a system. We were building disconnected tools.

Journey storyboard across multiple AI touchpoints

End-to-end journey storyboard, mapping how AI features could connect across the full member experience

Early ideation sketches across multiple AI surfaces

Ideation across surfaces: conversation patterns, AI disclosure moments, and cross-feature entry points

Key insight

The real opportunity wasn't better AI features. It was continuity. We needed to stop asking "where should we add AI?" and start asking "how might AI persist as a continuous partner across care?"

Phase 3: Defining the vision

Guide: a continuous care companion.

I led cross-functional workshops with Product, Clinical, Engineering, and Marketing to map the full member journey, identify where AI could carry context forward, and define system-level capabilities that no individual feature had addressed.

The resulting vision: a single AI-powered interface that persists across the entire care journey, carrying context, orchestrating capabilities, and meeting members wherever they are.

Vision framing: Guide's role in the member journey and the principles guiding its AI experience

Principle 01

Persistent memory

Guide carries context across sessions, knowing where a member has been and what they've already shared, so they never start from scratch.

Principle 02

Orchestration layer

A shared layer connecting all AI capabilities (provider fit, intake, summaries, support) into one coherent member experience.

Principle 03

Safety + trust built in

Clinician-augmented, not replaced. Every response meets VERA-MH standards for accuracy, compassion, and escalation to human support.

Care Guide spec, scheduling and navigation

Care Guide spec: scheduling flows and mobile navigation states defining the V1 member experience

Polished member experience, conversational care recommendation flow, H1 release

Contextual AI companion, accessible from every page for in-the-moment guidance and support

Defining Companion's surfacing logic, behavior states, and contextual entry points across the experience

Guide V2 shipped, H1 release deployed across 22 customer groups representing ~1.8M covered lives

In action

The conversation, end to end.

Screen recordings of the Care Guide conversational experience, members finding and booking care through AI guidance

Impact

From a bet to a roadmap.

Launched October 2025 across 22 customer groups representing approximately 1.8M covered lives, Care Guide validated that conversational AI can meaningfully reduce the friction that stops members from getting help.

The results gave us conviction to keep going. The 0→1 vision is now actively shaping Spring Health's AI roadmap. In 2026, continuous discovery is informing the next generation of Guide, expanding entry points, deepening personalization, and building toward a persistent companion that meets members wherever they are.

54% Of engaged members
booked an appointment

24K+ Appointments driven
through Care Guide

31% Of engaged members
started a conversation

8% Decrease in Care
Navigator calls

Booking conversion and engagement metrics from the A/B test, 22 customer groups, ~1.8M covered lives