About Proxify

Founded in 2018 and headquartered in Stockholm, Proxify helps 2,500+ companies worldwide scale their technical teams. The Financial Times ranked us as Sweden's 3rd fastest-growing company in 2025. Our clients increasingly need more than extra engineers. They need someone who can find where AI agents actually pay off and ship them into production.

The challenge

You're the AI engineer our clients have hired. You get close enough to their systems, data and processes to have a real opinion about what's worth automating and what isn't the right fit at the moment. You report to our Senior AI Product Manager and who owns discovery, commercials and scoping. The technical call is yours. Then you build it and lead Proxify engineers on the project, working within the client's existing data, integrations and security setup until it runs in production.

You'll sit in our core team with several client builds running in parallel. Around 80% of your week is hands-on building. The other 20% goes to keeping clients up to date and flagging what the product manager needs to know, as well as collaborating with them.

What you'll do

• Get under the hood of each client's architecture, data and integrations.

• Turn a scoped brief into an agentic system running in the client's own environment, end to end.

• Lead the engineers on each project and set the technical direction.

• Be the client's day-to-day technical contact: updates, questions, and the things that were clear on paper but not in practice.

• Keep several projects moving at once, and surface scope changes instead of agreeing to them on the spot.

Must-have

• 5+ years in software engineering, with a background in computer science, machine learning and generative AI.

• A track record of shipping LLM, RAG and agentic systems to production and keeping them running, with the depth to work out why a model behaves the way it does.

• Hands-on with RAG pipelines (vector databases, streaming, multi-tenancy), multiple LLM providers, and frameworks like LangChain, vLLM and PyTorch or TensorFlow.

• Strong Python, solid engineering habits (version control, testing, code review) and containerized deployments on AWS, GCP or Azure.

• Quick to pick up an unfamiliar codebase and build inside someone else's stack.

• A solution-based, consultative approach. You start from the client's problem, not from the tool or product you'd like to sell them.

• A strong communicator in professional English. You can lead a client conversation and explain technical decisions to people who'll never read the code.

• Self-directed. You run with a brief without needing someone to check in on you.

Nice to have

• Things you've built on your own time, such as side projects, open source or agent experiments, and the story of why you built them.

• Real experience with LangGraph, CrewAI, AutoGen or the OpenAI Agents API.

• Time at a startup or scale-up where priorities shift quickly.

Why you'll like it

You'll see a wide range of companies and problems in a short time, so the work never gets repetitive. There's no fixed playbook, so the ownership is real and the systems you build are yours to get right. You'll work directly with clients, with strong backing from leadership to move fast.

Practical details

• Location: Stockholm preferred

• Travel: occasional client site visits

• Interview process: Talent Acquisition call → hiring manager interview → practical AI engineering assessment → final interview