Build and own provider connectors on the new integration framework — auth, retries, webhooks, observability, etc.
Lead by example on technical direction — drive design discussions, unblock the team, and own the hardest pieces of the stack.
Partner with the client's PM and product stakeholders in 1–2 week sprints — daily standups, weekly demos — pushing back on scope, proposing better solutions, and translating product intent into clean technical bets.
Ship AI features end-to-end against the lighthouse roadmap (chatbot, Agent Co-Pilot, plumbing for what comes next).
Set engineering practices the permanent team will inherit — clean code, tests, AI-generated docs, CI/CD, eval harnesses on LLM-touching code paths.
Pair with the architect and tech lead on cross-cutting design; mentor through code review.
Requirements
6+ yrs building production back-ends; comfortable owning a service end-to-end
Strong in a modern typed stack (TypeScript/Node or Python/FastAPI)
Built and maintained third-party API integrations at scale (auth, retries, rate limits, webhooks, idempotency)
Hands-on with agentic AI tooling — Claude Code in daily SDLC, LLM API integration, prompt + eval discipline
Opinionated and direct — pushes back on scope, holds the bar on docs and tests
Cloud + IaC (AWS or GCP, Terraform, Docker, CI/CD pipelines)
**Bonus: **VLM/RPA or browser-automation experience for provider fallback paths
Tech Stack
AWS
Cloud
Docker
Google Cloud Platform
Node.js
Python
RPA
SDLC
Terraform
TypeScript
Benefits
You'll be working alongside a talented, hand-picked senior team.
Greenfield rebuild, not legacy maintenance. We're migrating off the current vendor's platform onto a new, clean, modern codebase. You'll set the future patterns the new system gets built on.
Best-in-class agentic-AI SDLC from day one. You'll help design the patterns — Claude Code across the full SDLC, AI-generated documentation, evals on every LLM-touching feature, observability built in.
Interesting consumer AI features — a live-shopper dashboard, an in-funnel AI chatbot, and the plumbing for everything that comes after.
Path to lead the team going forward. This 3-month build is Phase 1. The client will then ramp a permanent squad on top of the codebase you helped create — clear path into the lead role.