Product Ops, rewired around Claude Code.

The challenge

Modern Health is the leading global mental health benefits platform, trusted by employers like Lyft, EA, Ancestry, and The Trade Desk to support their people. The Product Operations team owns the shared evidence layer that feeds Product and Tech planning across the company: VOC tickets, Salesforce signals, Gong calls, QBR notes, the Opportunity Registry, the XFN forum cadence. Sweeps were manual, signals got stuck in their first channel, and the team's facilitation hours were leaking into prep and follow-up work that AI could carry.

What we did

We ran a one-day on-site workshop in San Francisco built around Claude Code and the modern MCP stack. The room was split into three concurrent tracks, each owned by a real leader on a real artefact: a VOC-to-opportunity synthesis track on Modern Health's own ticket data, a brand and design-system track that built a working HTML template against their branded deck and Figma library, and an accessibility-and-research track that audited real Figma frames and turned a de-identified Zoom transcript into structured research notes.

Each track left the room with a working prototype against their own data, not a sandbox. The artefacts pointed straight at the four high-leverage AI use cases we identified during the pre-engagement audit: an always-on Opportunity Registry scoring agent, an XFN forum co-pilot, a WPO / Care Connect operations co-pilot, and a Product Ops research synthesis loop.

The outcome

The team left with the patterns, the tooling, and the language to keep building. Every track artefact carried straight back into the existing workflow stack (Notion, Jira, Salesforce, Glean, Gong, Slack) with Claude Code as the new connective layer. The engagement continues.