Our mission: eliminating every barrier to mental health.

Spring Health is a global mental health company on a mission to eliminate every barrier to mental health. We're building a world where getting support is simple, personal, and built around the person, so care can continue through every job, move, health plan, and life stage.

Our AI-native platform helps us deliver personalized support across self-guided tools, coaching, therapy, medication management, and specialty care. With outcomes independently validated by JAMA Network Open and the Validation Institute, Spring Health reaches more than 170 million people worldwide through leading employers, health plans, and partners.

As an AI-native company, we believe technology should expand the reach, quality, and humanity of care. Every Spring Health team member is expected to use AI tools thoughtfully, apply human judgment to AI outputs, and keep building AI fluency in ways that support their role and our mission.

Reporting to the Senior Director of Care Operations Intelligence, this Senior AI Care Operations Manager is a hands-on individual contributor who builds, tests, and improves the workflows behind Spring Health's support experience, for AI agents and for people. The role has two parts: building and improving AI-agent workflows, including agent operating procedures (AOPs), and engineering better human and automated processes from current-state mapping through measured results. This is a full-time, remote role, with a preference for being onsite in NYC once a week. Travel is minimal, at most once a quarter.

What you’ll do

Build and own AOPs. Turn policy, workflow, and edge cases into decision logic that AI agents follow reliably, across Spring Health and Alma. Own each AOP's results after launch, and keep it current as products and policies change, working with Product and Engineering before releases.

Test AOPs and investigate failures. Test AOPs before launch with simulations and representative interactions, and help set the standard for "ready to ship." When AI fails, find the root cause, fix owned AOP issues, and document and route the rest to the right owner. This is separate from the independent QA function, but partners with it.

Map how work actually gets done. Map current workflows and member and provider journeys, including decisions, handoffs, exceptions, controls, and sources of delay or rework.

Design the future state. Quantify the problem with frontline staff and operational data, then design a simpler process with a clear owner and a measurable outcome. Look beyond Decagon for automation, and for ways to improve the underlying human process.

Build, pilot, and measure. Develop the business case and prototype or configure within approved tools, partnering with Product, Engineering, CX Tooling, Platform S&O, and frontline leaders on builds and releases they own. Pilot changes, support rollout with Care Delivery Enablement, and measure quality, speed, experience, and cost.

Work within clear boundaries. Knowledge Management owns knowledge content, and Experience Intelligence & Quality owns QA methodology and scoring. Use its signals to diagnose problems. Frontline leaders own operational policy, human-team performance, and adoption decisions.

What success looks like in this role

AOP delivery and reliability: Priority support intents covered by tested AOPs, simulation results, post-release defects or regressions, and resolution quality, escalations, handoffs, and repeat contact for owned AOPs. Specific KPIs: Decagon eligibility rate, Decagon resolution rate, and Decagon CSAT.

Failure response: Time to reproduce and diagnose a high-priority issue, quality of the root-cause write-up, time to implement or route a fix, and recurrence after remediation.

Process visibility and redesign: Priority workflows mapped with baseline measures and agreed future states, with fewer unnecessary steps, handoffs, delays, and sources of rework.

Implemented improvements: Redesigned workflows in use with sustained adoption, and measured change in cycle time, error rate, repeat work, access, and member and provider experience.