HFM is an internationally acclaimed multi-asset broker, delivering cutting-edge trading tools, platforms, and conditions to traders worldwide. We are committed to innovation, transparency, and excellence in the financial markets. As we continue to expand, we are seeking a driven and strategic Mid AI Engineer to join our team.
Responsibilities:
- Build and ship internal AI services end to end — from the data pipeline through the model to the API other teams call — working within the team's established architecture.
- Develop asynchronous REST APIs and microservices in FastAPI, deployed on Kubernetes, to serve models and orchestrate agent tool calls.
- Implement retrieval pipelines over PostgreSQL and pgvector — chunking, embedding, indexing and retrieval tuning — and measure them against a held-out evaluation set.
- Build and maintain training pipelines covering data preparation, fine-tuning, evaluation and model registration, promoting a new model only when it measurably beats the one in production.
- Integrate agent tools and Model Context Protocol (MCP) endpoints, ensuring each tool enforces its own access controls rather than relying on the model to do so.
- Write the evaluation sets, structural output checks and drift monitoring that prove a model works before release and keep proving it afterwards.
- Instrument services with structured logging, Prometheus metrics and OpenTelemetry tracing, so that failures can be diagnosed rather than guessed at.
- Review the work of junior engineers and trainees, pair with them on delivery, and contribute to the team's engineering and documentation standards.
- Requirements Bachelor's degree in Computer Science, Artificial Intelligence, Software Engineering, or a related field.
- 3+ years of backend software development experience, including hands-on delivery of at least one machine learning or LLM application into production.
- Strong proficiency in Python and FastAPI, with solid working knowledge of PostgreSQL and SQL.
- Practical experience with LLM application patterns — retrieval-augmented generation, prompt design, structured outputs, and evaluation of model quality.
- Experience with Docker, Git and CI/CD pipelines, and comfort working in a containerised environment such as Kubernetes.
- Working knowledge of a major cloud platform, ideally AWS.
- A methodical approach to correctness: you write tests, you measure before and after, and you can explain how you know a change is an improvement.
- Strong communication skills and full professional proficiency in English.
- A collaborative approach — comfortable learning from senior colleagues and supporting those with less experience.
Nice to have:
- Experience running vector search in production (pgvector, OpenSearch, Qdrant or similar).
- Exposure to multi-agent frameworks (e.g., LangGraph, CrewAI, AutoGen, LlamaIndex) or to the Model Context Protocol.
- Experience with model registries and deployment workflows (MLflow or equivalent), or with serving open-weight models (e.g., vLLM).
- Familiarity with analytical data stores such as ClickHouse, and with observability tooling (OpenTelemetry, Grafana, Langfuse).
- Familiarity with FinTech platforms, FX trading, or financial services operations.
- Resumes must be submitted in English.
- Applicants must be eligible or have legal authorization to work in the country where the position is based.
- Benefits Hybrid Work Model (2 days working from home) Monthly Food Vouchers Comprehensive Health & Life Insurance Summer Short Fridays (August) Additional Paid Annual Leave Birthday Leave Udemy Business access Referral Bonus Program Additional Support Visa Sponsorship and Relocation Assistance Sounds like you?
- Let’s write the next chapter together!
- All applications will be handled with the strictest confidentiality.
