HE3 - Software Developer, E-Commerce Systems

We run three home furnishings storefronts on a catalog of 100,000+ SKUs from a long list of manufacturers. Almost all of it runs on software we built ourselves: a Django backend, a React storefront on Cloudflare Workers, and a set of pipelines that turn messy vendor spreadsheets, emails, SFTP drops and websites into clean product, pricing and inventory data.

We're looking for a developer to work alongside our lead developer on that stack. Most of the job is solving concrete business problems: a vendor feed that breaks every month, a pricing rule nobody can explain, a task the team still does by hand in a spreadsheet. You'll figure out what's actually going on, then build the fix, and look for places where AI can make the process itself work better.

What you'll work on

• Product data. Vendor imports and feeds, our in-house PIM, categorization, images and data quality across a large, inconsistent catalog.
• Inventory and pricing. Feed ingestion (AWS Lambda, S3, SFTP), scrapers for vendors without feeds, and the pricing, deals and cart logic that depend on them.
• The storefront API. We're replacing Shopify with our own Django backend and a React Router storefront. You'll build and test the API endpoints the new site depends on.
• Search. Elasticsearch-backed search, facets and recommendations.
• Internal tools and automation. Find the manual, repetitive work across the business (merchandising, customer service, operations) and replace it with something reliable. Sometimes that's a script or an admin screen. Sometimes it's an LLM-based workflow, if that's the right tool for the problem.
• Infrastructure. We're moving hand-built AWS and Cloudflare resources to OpenTofu. Some exposure here is useful but not required.

How we work

• Tests come first. We write a failing test, then the code. Tests build their own data and assert on exact values. A test that only checks for a 200 response doesn't count.
• Typed, reviewed, small. Python is type-annotated and checked, TypeScript runs in strict mode, and changes go in as small, focused pull requests with one fix per commit.
• Production is careful work. Migrations, data fixes and infrastructure changes get agreed on before they run, and we don't commit secrets.
• We use AI coding assistants every day, and we review everything they produce. We keep written project conventions that the agents follow. We want someone who gets real speed out of these tools and still owns every line that ships, and who can spot when the assistant got it wrong.

What we're looking for

• Strong Python. Django (ORM, migrations, admin) experience is a big plus.
• Solid SQL and PostgreSQL. Comfortable reading query plans and pushing work into the database when it makes sense.
• Experience building and consuming APIs. Django Ninja, Pydantic or OpenAPI is a plus.
• Experience working with messy real-world data: CSV/Excel imports, vendor feeds, scraping, ETL.
• Enough TypeScript/React to work on the storefront when a feature spans both repos.
• Good judgment about when automation (including LLMs) helps and when a plain script or a better process is the right answer.
• Clear written communication. You'll write PR descriptions, docs and tickets that someone else can act on.

Nice to have

• E-commerce experience, especially catalog, PIM, order management or checkout (Stripe).
• Elasticsearch or another major search platform.
• AWS (Lambda, S3, ECS), Cloudflare Workers, OpenTofu/Terraform.
• Experience in furniture, home furnishings or another product-heavy retail business.