Full-Stack Engineer

Why this role exists

Our client helps hospitals identify purchased-services overcharges before payment. Contracts and invoices move through extraction, validation, retrieval, and LLM-assisted reasoning; the product turns that work into defensible, line-level findings and recommendations for accounts-payable analysts. The company is expanding its client base and moving its product from MVP-era implementation toward reliable, scalable enterprise operation.

About the Role

This is an early, high-ownership engineering role responsible for the product surface analysts use and the system seams that feed it. You will ship across a production TypeScript stack, trace defects across the app, sync, and validation pipeline, and improve the reliability and performance of a multi-tenant enterprise product.

The first priority is strong full-stack engineering: delivering complete features, protecting data meaning, and owning what reaches production. You will also work with SpendRule’s agent layer and AI-assisted development environment, applying the same verification standard to generated output as to any other code.

What You’ll Own

● Ship analyst-facing features end to end, from Prisma/Postgres models and migrations through Express routes, typed clients, and React/Next.js experiences you verify in the browser.

● Build for incomplete and delayed data. Treat loading, empty, partial, and error states as part of the feature, and preserve the difference between unknown, zero, verdict, and recommendation.

● Diagnose issues at the seams between the app, sync jobs, and validation pipeline. Use queries, logs, traces, tests, and reproduction evidence to find the real failure point.

● Improve performance and reliability across large data grids, queries, routes, tenant-scoped workflows, auth gates, rate limits, audit logging, and internal administration tools.

● Use AI coding agents deliberately: provide precise plans, review generated diffs line by line, validate edge cases, and retain full ownership of the result.

● Close work with an honest evidence trail: root cause, what changed, how it was verified, what remains untested, and which follow-up risks still exist.

What Success Looks Like

First 30 Days

● Understand the path from contract and invoice ingestion through validation, sync, and the analyst-facing application.

● Run the product and test suite, learn tenant and data-integrity boundaries, and map the highest-risk portal seams and performance constraints.

● Deliver an initial production-ready improvement with clear verification evidence and appropriate code-review support.

By 60 Days

● Ship at least one meaningful customer-facing feature or improvement across the full stack.

● Diagnose and resolve an issue that crosses application or system boundaries.

● Demonstrate effective use of the team's AI development tools while independently reviewing and validating their output.

● Contribute improvements to reliability, performance, maintainability, or developer workflow.

By 90 Days

● Independently own and ship a meaningful full-stack portal feature, including partial and failure states, tests, rollout, and monitoring.

● Diagnose and improve at least one material latency, reliability, or data-integrity issue across multiple system layers.

● Take dependable ownership of portal work so current technical leaders can return more capacity to core back-end product priorities.

● Demonstrate disciplined use of AI-assisted engineering through clear plans, rigorous review, and evidence-backed verification.

Qualifications

● 5+ years shipping production TypeScript across both sides of the wire, including React or Next.js and a Node back end with a relational schema.

● Strong SQL fundamentals and practical Postgres experience, including joins, indexes, query plans, and performance diagnosis.

● Evidence of owning customer-facing features through design, implementation, testing, rollout, and production follow-through.

● Experience debugging across multiple services or layers and measuring before-and-after results instead of relying on intuition.

● Experience using AI coding agents on real work, with concrete examples of directing, reviewing, verifying, and correcting their output.

● Experience building operational interfaces such as tables, filters, detail views, document viewers, exports, or other dense workflows, with attention to accessibility and incomplete states.

● Ability to operate from incomplete specifications, ask focused questions, make appropriate decisions, and communicate technical findings clearly to non-technical stakeholders.

Strong Preferences

● Experience with Prisma, Express, Next.js App Router, multi-tenant SaaS, enterprise auth, Redis, S3, or comparable technologies.

● Performance work on large grids, high-volume queries, or sync-driven applications with measured results.

● Experience in healthcare, procurement, AP automation, fintech, audit-heavy workflows, or another domain where data claims must be defensible.

Nice to Have

● Python or FastAPI exposure and the ability to read adjacent pipeline code.

● Experience with RAG, vector search, LLM evaluations, tool-calling agents, or customer-facing AI output.

● Experience with SAML/OIDC, SCIM, tenant provisioning, or enterprise access controls.

Compensation & Benefits

• Competitive compensation
• Significant ownership and influence as part of a small engineering team
• Opportunity to work at the intersection of applied AI, enterprise software, and healthcare financial operations