Founding AI Backend Engineer (Contract, path to full-time)

Founding AI Backend Engineer — Aistetic

Aistetic builds AI-powered listing tools for fashion resellers, turning product photos into ready-to-publish listings for eBay, Vinted, and Depop. We're an Oxford University spinout, backed through Techstars and eBay's Future of Ecommerce Accelerator, with 600K+ items processed through our AI engine.

The Role

Aistetic's core product runs on a 30+ agent AI architecture — a multi-stage pipeline combining computer vision, LLM inference, and structured output generation, deployed on a fully containerised, Terraform-managed cloud infrastructure. You'll be the technical owner of this stack: the AI engine, the cloud infrastructure, and the marketplace integration architecture — including how we orchestrate around the company's central technical risk, Vinted's lack of a public API, working jointly with the mobile engineer on the client-side automation itself.

What you'll do

• Own the AI/ML pipeline: 30+ agent architecture spanning computer vision, LLM inference, and structured listing generation
• Own infrastructure end-to-end: Kubernetes (currently v1.30), Terraform IaC across dev/staging/production, GPU inference nodes (NVIDIA L4)
• Support and own the cloud setup configuration, maintaining uptime and customer delivery Own marketplace integration architecture — current (eBay, Vinted) and roadmap (Poshmark, Depop, Facebook) — including orchestration, monitoring, and retry/failure handling for how publish events flow through the backend
• Share the build-vs-risk call on Vinted with the mobile engineer, who owns the client-side non-extension automation itself — you own the architecture and risk framing around it, not the automation build
• Take structured handover from the outgoing backend engineer before he's gone Work closely with the mobile and front-end engineers on shared architecture, API contracts, and integration points — this isn't a solo silo

Technical Requirements

• Strong backend engineering with production Kubernetes experience (node pools, autoscaling, private clusters)
• Infrastructure-as-Code fluency — Terraform specifically, managing multi-environment deployments (dev/staging/prod)
• Experience with cloud-native data layers: managed PostgreSQL, Redis, message queues (RabbitMQ or equivalent)
• AI/ML pipeline experience — LLM inference serving, GPU workload orchestration, ideally computer vision exposure too
• Comfortable working with undocumented or unofficial third-party APIs — directly relevant to Vinted
• Cloud migration experience across cloud providers — comfortable reading and translating provider-specific infrastructure into equivalent managed services elsewhere
• Security fundamentals: workload identity, secrets management, network isolation in a Kubernetes context

Ownership & Judgment

• Comfortable owning the highest-stakes infrastructure and architecture decisions, with no senior technical peer in-house to check against
• Sound risk/reward judgment on ToS-adjacent integration work (Vinted), held jointly with the mobile engineer — knowing when to push, patch, or pull back on what gets built
• Systems thinker: this role spans infra, AI/ML, and integration architecture, not one narrow lane
• Founder-adjacent mindset: comfortable working directly with a non-technical CEO, translating technical trade-offs into plain terms

Product & Collaboration

• Obsessive about product and customers — instinctively looking for how to deliver value quickly and effectively, not just technically correct
• Brings ideas to product vision and roadmap — doesn't need to own it, but actively shapes it rather than just executing specs handed down
• Works closely with mobile and front-end engineers day to day, not in isolation
• Strong communication skills — collaborative, fast, and efficient, able to translate technical trade-offs clearly to both teammates and a non-technical founder

Structure

• 3–4 days/week to start, with a path to full-time
• Equity included as part of the package post initial phase, terms to be discussed
• Comp: Monthly Fee dependent on candidate