Skip to content
Projects & case studies
App Store deployed · Preparing for commercial launch

Phase 3 Benefits

Eight benefit accounts on one card.

Health, commuter and wage-law benefits on a single payment card, with a member app on web, iOS and Android and an admin portal. Behind them, the systems one person uses to run product, engineering and compliance for a regulated financial product.

A Phase 3 Benefits payment card on a paper background.

What it is

Who it is for
Members who carry the card and use the benefits on it, and the program team and partners who operate it.
What it covers
Eight health, commuter and wage-law account types on one payment card. A member app on web, iOS and Android for onboarding, the card and wallet, balances, receipts and claims. An admin portal for the operators' side of the same product.
What runs behind it
A second brain that holds requirements, decisions, schedule, budget and the answers given to partners, and rewrites itself when something changes. A SysAdmin Toolkit through which agents provision and run the infrastructure, with SOC 2 and HIPAA policies and evidence grounded in what was implemented.
How it is run
One person directing AI agents across product, engineering and compliance. The second brain keeps everyone, human or agent, working from the same facts. The toolkit gives agents one way to operate the infrastructure.

Why the way it is run is worth reading about

Most writing about AI in product work is about producing things faster. The interesting part of Phase 3 Benefits is what it takes to keep a regulated financial product true while it changes weekly.

A benefits product with a sponsor bank behind it accumulates commitments in meetings, and several of them depend on what a partner's API can actually do. The infrastructure it runs on accumulates changes just as quickly, and a regulated product has to be able to describe them. The two systems behind the product exist so that neither accumulation gets ahead of the record: the plan is updated from the last conversation, and the policies are grounded in the environment as it was actually built.

Two of these are what members and operators see. The other two are the reason one person can run them, and each has a case study of its own.

Case studies

Short, specific, and each one built around a single question a product or engineering leader would actually ask.

Need a leader who can
build with AI?

Let's talk about the product challenge, the team, and the outcomes you need—whether you're improving how your business runs or building something new.