SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Gigi is an agentic operating system for enterprise media buying, with an AI media manager for Amazon DSP launched in summer 2025. The company has grown to manage hundreds of millions in advertising spend for major agencies and is backed by Golden Ventures and Aperiam Ventures. The founding team brings 15+ years of Amazon Ads and ecommerce experience, including a prior $100M+ exit. Revenue has doubled in the past two months with expectations to double again by early Q4.
You will join as the 12th employee, a full stack software engineer owning end-to-end feature development for Gigi's AI agent. This is fundamentally a software engineering role combining familiar work—designing APIs, building interfaces, modeling data, integrating services, writing tests, debugging production issues, and making architectural decisions—with agentic systems where LLMs reason, use tools, and take actions on behalf of customers.
You'll ship features from the data layer through agent logic to the customer-facing interface. The tech stack includes Python, FastAPI, LangGraph, and LangChain on the backend; TypeScript, Vue, and Nuxt on the frontend; Java and Spring Boot for services; PostgreSQL, pgvector, Redis, and NATS for data; and AWS infrastructure. You'll use AI coding tools like Codex, Claude Code, and Cursor as core parts of your workflow while remaining accountable for architecture, quality, and correctness.
Key responsibilities include building agent judgment (deciding what Gigi should do autonomously versus escalate to humans), drawing lines between probabilistic and deterministic work, designing reliable agentic systems with tools, retrieval, memory, evaluations, guardrails, and approval flows, and working directly with customers to turn ambiguous problems into shipped product.
Success milestones: code in production within 30 days with understanding of media buyer workflows; owning a product surface area by six months; shipping something unique and iterating on abstractions by twelve months.
You should have a track record of building and shipping production software, strong full stack fundamentals (system design, APIs, data models, interfaces, testing, debugging), comfort across the stack with real depth somewhere, and excitement about using AI coding tools daily. Curiosity about how LLMs change software products and the ability to make reasoned calls and iterate fast are essential. Prior AI research, LLM production experience, advertising domain knowledge, or early-stage startup experience are bonuses but not required.
