Is Spring Health Building a Mental Health Operating System?

I reviewed Spring Health's Q2 product webinar, and at the heart of the system lies a deep issue: context.

When patients move from therapist to therapist, all context is lost between sessions. Reintroducing and rehashing their case is not only mentally taxing for the patient, it can be dangerous in moments of crisis.

Three launches stood out to me:

1. Guide. An AI agent that helps patients at every step of the journey. A standout moment in the demo was Guide offering the patient a breathing exercise in the middle of a session, right when it was needed.

2. Work-Life Services. A lot of the stress and anxiety patients carry comes from upstream issues: legal problems, finances, immigration. Work-Life Services connects members to local resources for each of these directly in their account. I find this especially interesting because it expands the gamut of what counts as mental health care.

3. Pre-Session Review. Therapists see a summary of patient notes across multiple dimensions before each visit, which means less time re-establishing context and more time delivering personalized care.

Looking at these together, I realized Spring Health's Q2 launch looks less like a set of features and more like an operating system:

1. Memory. Patient summaries carry context to every therapist or mental health professional the member reaches out to.

2. Applications. Therapy, coaching, meditation, and work-life services act as programs the member can access, all reading from the same memory.

3. Scheduler. Guide decides what deserves the member's attention today.

Features compete one by one. An operating system competes on the layer underneath, and that layer gets more valuable with every session.

The open question is how you measure success for Guide. If the North Star is engagement, the system learns to keep members talking to it. But in mental health care, success often means a member needing less support over time, not more. That is a rare product where the best outcome can look like declining usage.

Having spent nine years on the payer side of health care, I would add one more wrinkle. There is no claim code for "AI guided a member through a breathing exercise between sessions." If Spring Health can connect Guide activity to measurable clinical outcomes, they are not just building a better product. They are building the evidence base that lets employers, and eventually payers, pay for what happens between sessions instead of just the sessions themselves.

If you work on any of these products at Spring Health, I would genuinely love to hear how you think about this. Which metric would you protect at all costs?