Senior AI Engineering Lead

Why This Role Exists

Nobody on our team writes code by hand. Claude does it. The engineering job is architecture, review, and keeping the system coherent. Understanding what AI produces, making sure new pieces integrate with what's already there. Not typing code. Thinking about systems.

We need engineer #1 who works this way. You steer the AI, review what it produces, iterate until it's right. By end of day you've shipped a new customer integration that fits cleanly into the existing architecture. And you know how it works.

About the Workflow

We build with Claude Code. AI writes the code, you review it, you keep the system coherent. The product itself integrates four LLMs (ChatGPT, Gemini, Claude, Perplexity), so you'll work across all of them.

Async, AI-augmented collaboration: you ask a question, it gets routed through an agent, you get back a markdown file. That's how we communicate.

If that sounds like how you already work, keep reading.

About AirShelf

AirShelf AI is the agentic commerce platform. We transform product catalogs into AI-ready formats so brands get discovered and purchased by AI agents across ChatGPT, Google, Perplexity, and emerging platforms.

Backed by Antler Singapore. Part of HP Garage 2.0. Enterprise clients in APAC. Our cofounders built and scaled commerce platforms serving millions of users across APAC — they didn't just ride the e-commerce wave, they helped create it over the past 15 years.

What You'd Own

- Customer-facing dashboard (product golden records, AEO insights, score tracking)

- Data pipelines (product enrichment, benchmarking across AI models)

- CLI tools (agent readiness scoring, leaderboard)

- New customer integrations end-to-end

Tech Stack

- Next.js, React, TypeScript (frontend)

- Python (pipelines, enrichment, scoring)

- ChatGPT, Gemini, Claude, Perplexity — four LLMs, all first-class in the product

- Agentic tools (MCP, function calling, tool use)

- PostgreSQL, Vercel, PostHog, Axiom

- Claude Code as primary dev workflow (Codex, Cursor also fine)

What We're Looking For

- Strong engineering fundamentals — you need to understand what good code looks like to catch when AI-generated code isn't

- Systems thinking — how components fit together, where abstractions leak, what breaks at scale

- Comfortable NOT writing code yourself — your job is directing, reviewing, and maintaining coherence

- Can explain how something works after AI built it — "I don't know, AI did it" is a failure mode here

- Ships and integrates — code that works in isolation but breaks the system doesn't count

- Good async communication — remote team, markdown over meetings

What We Don't Care About

- Years of experience on a resume — we care about judgment and systems sense

- Which specific AI tools you use — we care that you use them as infrastructure, not novelty

- Computer science degree (but strong CS fundamentals from wherever you got them)

What You Get

- Engineer #1 — you shape the system, not inherit someone else's mess

- Meaningful founding equity (early stage)

- Work directly with the CTO — daily guidance on architecture, no layers

- Full remote, async-first

- Small team: 2 cofounders (CEO + CTO), designer, product advisor

- Direct access to enterprise customers from day one

Hiring Process

Fast, human, and real.

1. Intro Call — Chemistry and alignment. 30 min with the CTO.

2. Live Build + Design Discussion — Build something real with AI on a shared screen. Steer the AI, review its output, iterate. Then discuss the architecture decisions. We want to see how you think, not what you memorized.

Reach out: talent@airshelf.ai