Senior AI Engineer (Python, Django, Angular, RAG)

Brief introduction
You will build production RAG and LLM systems that read real documents and run inside enterprise ERP and finance stacks.
This is a solo, end-to-end role. You join client calls, design the architecture, then build the important parts yourself. Our core stack is Python, Django, Angular and RAG. Ukraine/EU hours.

About us
Deploy Staff automates the manual work large enterprises run on: finance, operations, customer service, reporting. We build complex systems for clients in the US and UAE (currently primarily UAE), centered on production RAG and LLM workflows that read real documents and integrate with real ERP and finance systems.
Despite the name, we don't place engineers into other people's teams. We are not an outstaff company. We own the solutions we build end to end, and so will you.

The role
Solo and end to end. You are the engineer on your projects: there is no development team, and the technical work is yours from discovery through production support.
You sit in on client calls with the project manager, understand the business problem, design the architecture, then open the codebase and build it yourself. The PM runs the relationship and the commercial side. You own the system.
Solo does not mean isolated. You have direct access to the people making decisions, and we talk architecture through with you before you commit to it. After that the code is yours to write and merge. Part of the job is judgement about which calls are yours and which should come back to us, and we would rather be asked than surprised.
If you have spent recent years distributing work and reviewing it, this will feel like a step backwards. If you have spent them wishing there were fewer people between you and the problem, it will not.

What we expect
Our core stack is Python, Django, Angular and RAG/LLM-based workflows. That is what we need you to already be good at.
Core
Production LLM work: RAG (retrieval design, chunking, evaluation), document processing, structured extraction or agentic workflows. Shipped and operated by you, through the parts that only show up after real users arrive
Strong Python with Django or DRF, and PostgreSQL data models you can design and debug in production
Angular and TypeScript. You own the frontend of what you build, so you need to be able to ship a feature through it without waiting for anyone
Advanced English (C1). You will question senior stakeholders directly and explain a decision a second way when the first one does not land
Also part of the work
You will hit all of this. We do not expect equal depth in every line, and we do expect you to pick up what you are missing.
Celery, Redis, queues and asynchronous processing
Enterprise integration: ERP, finance, CRM systems and their APIs
Cloud infrastructure, deployment, observability and security. AWS, GCP, Kubernetes, Docker
Reading an existing system before changing it, and keeping what already works

What matters more than the CV
What you personally built, what you personally decided, and what happened when it broke. If your strongest examples are systems your team implemented while you supervised, this is not the right fit.

How we work
Remote, with working hours aligned to the Ukraine time zone.
Small team, direct access to decision makers, real operational problems. You get problems to solve rather than specs to implement.

Interview
1. Introductory interview.
2. Technical interview with us.
3. Test task.
4. Offer.
Interviews are recorded and shared only within the hiring team, so decisions are made on what you actually said rather than someone's notes.

Repeating once again - our core stack is Python, Django, Angular and RAG! We cannot compromise on any of these. If one of those isn't present or barely used - this is an immediate mismatch!