Software Engineer

Revin•New York, NY

23d•$140,000 - $180,000•Onsite

About The Position

We're hiring a Software Engineer to build the production systems behind our AI agents — the agents that handle SMS, voice, and email conversations on behalf of $100M+ home-services operators. When our agent makes a mistake — books outside hours, dispatches to the wrong zip code, promises a service the business doesn't offer — it costs our customers real money. Your job is to help build the systems that prevent that. This is a high-ownership role. You'll work end-to-end across a Python/FastAPI backend, a Next.js/TypeScript dashboard, and the LangGraph workflows that drive agent behavior. You'll partner directly with product, design, and senior engineers, ship in production, and grow quickly.

Requirements

2–4 years of full-stack experience shipping production software at a startup, high-velocity SaaS, or AI-driven company
Eager to work with LLMs in production — you've built with LLM-powered features and want to go deeper, even if you haven't owned them at scale yet
Working proficiency in Python (FastAPI) and TypeScript (Next.js, React Query)
Comfortable working in production systems, and eager to learn multi-tenant patterns — RLS, account-scoped queries, JWT-based access
Experience debugging APIs and tracing issues across services and data pipelines
A continuous learner comfortable operating in ambiguity. You ship, learn, and adapt.
Customer-obsessed — you want to be close to the people using what you build

Nice To Haves

LangGraph, LangChain, or other agent frameworks
Real-time voice/audio work (LiveKit, Daily.co, Twilio Voice)
Observability tooling for LLM systems (Langfuse, LangSmith, Helicone)

Responsibilities

Build and evolve LangGraph workflows that handle SMS and voice conversations for enterprise operators
Ship features end-to-end across a Python/FastAPI backend and a Next.js/TypeScript dashboard
Work in a multi-tenant Supabase/Postgres environment where every query is account-scoped (RLS, JWT-based access)
Help build the eval frameworks that catch agent regressions before they reach customers — we treat eval quality as a first-class engineering discipline
Integrate with messy real-world CRMs (Salesforce, ServiceTitan, LeadPerfection, AccuLynx) where the data is never clean
Partner with the voice team on real-time agent behavior (LiveKit + OpenAI Realtime)
Help optimize the omnichannel experience across SMS, voice, and email