The company uses technology to help people get through their employer-sponsored health plans, aiming to lower costs for employees and employers without lowering the quality of care.
It is building an agent-driven navigation platform for self-insured health plans, aimed at the main driver of rising healthcare costs.
The team is small and moves fast, with people based in NYC and New Orleans.
The Platform team owns the member experience, member management, the agent harness, the channels members use, and the infrastructure under them.
The team values a very high bar delivered with humanity, full ownership of problems from start to finish, direct and early communication, small batches shipped quickly, and results over meetings.
The work involves complex health plans, regulated data, and new agent capabilities, and the product is being shaped as the category is rebuilt around agents.
Members and admins use what you build, so you will hear from them directly rather than through layers of tickets.
Responsibilities
Join the Platform team as a senior engineer.
Build the system that answers a parent's urgent late-night question within seconds: where to take a feverish child when the pediatrician is closed, and what it will cost. This uses protected health data, dependent coverage, the family's deductible, and the urgent care and telehealth options in the plan, and the answer must be right every time.
Own the architecture, trade-offs, and results for the systems you build, and help decide where they go next.
Build member features such as insurance cards, claims, accumulators, dependents, and provider search.
Model benefits and entitlements precisely so the app and the agent give the same correct answer to a member.
Build admin features for member and company management, including onboarding, benefits configuration, and usage analytics.
Build internal tools so the team can configure and support customers without an engineer.
Ship features across web, iOS, Android, and the browser extension.
Agree requirements and trade-offs with product and design before building, and talk with members and admins to see how the product is really used.
Build the agent tools, entitlement data, and plan data that let the member agent answer from what each member actually has, working closely with the engineers on the agent harness.
Build the channels the agent uses (web, mobile, SMS, Slack, Teams, and MCP clients) so it behaves the same in each.
Add evals and tracing to what you ship so wrong answers are caught before members find them.
Own parts of the AWS infrastructure, CI/CD, and the pipeline from development to production.
Improve observability through structured logs, metrics, traces, alerting, and incident response.
Build security into the architecture: minimize PII, use envelope encryption with key rotation, enforce least-privilege access, keep purpose-of-use audit trails, and support break-glass workflows.
Improve developer tooling, including agent rules and skills, linting, the CLI, and automated tests, so engineers and agents ship faster and more safely.
Turn better practices into rules, skills, or checks so the whole team and every agent follows them.
Bring significant trade-offs to the team early, with a recommendation.
Lead incident response for the systems you own.
Within your first 90 days, own two or three platform systems and be the first contact when they fail.
Within your first 90 days, ship at least one member-facing feature end to end, from requirements through production monitoring.
Within your first 90 days, close one gap in observability, CI, or developer tooling that speeds up the whole team.
Stack: TypeScript, Next.js, Vite, Hono, PostgreSQL, Temporal, Capacitor for web, iOS, and Android, AWS with Terraform, Sentry, and Amplitude.
Requirements
Have 4+ years of production software experience, or have built equivalent work in less time. Results matter more than years or employer names.
Have shipped complex, user-facing systems end to end, designing them from the data model through the API to the UI.
Understand software design well: you model domains carefully, draw clean boundaries, plan for failure, and change systems without breaking dependents.
Have spent at least a year shipping production code mainly with coding agents, and built the rules, skills, tests, and evals that keep their output correct.
Have built on regulated or sensitive data, with access control, audit trails, and least privilege designed in from the start.
Are an expert in TypeScript.
Communicate clearly in writing and in conversation, and can take a trade-off from discussion to decision with engineers, designers, or customers.
Bonus: you have built LLM-powered product features, including agent tool design, context engineering, evals, guardrails, and trace interpretation.
Bonus: you have worked on multi-tenant SaaS, including tenancy, white-labeling, and SSO (OIDC, SAML).
Bonus: you have used durable execution or workflow engines such as Temporal, or built event-driven systems.
Bonus: you have shipped across web, iOS, Android, browser extensions, and internal admin tools.
Bonus: you have worked in healthcare, benefits, insurance, or payroll, including FHIR, EDI, or HL7.
Bonus: you have built or maintained HIPAA- or SOC 2-compliant systems.