#vacancy #backend #python #ai #llm #agentic #ops #europe #job #remote
Agentic Systems Engineer (Python)
Company: A startup building a next-generation donation service. It creates public ratings of content and authors (movies, series, books), where donations are converted into protected markers of quality and reputation.
Location: Remote (Europe / CIS, excluding Russian Federation and Republic of Belarus).
Essence of the role:
We need an engineer who uses agents as their primary work tool every day. Someone who can take a vague business task and turn it into a working system, managing agents like a team: assigning tasks, controlling results, iterating.
Responsibilities:
- Design and build multi-step agentic workflows as the working layer of business infrastructure: with memory, tool calling, human approval at necessary points, and quality control.
- Translate existing rating developments into a stable, automated, and scalable production system.
- Independently understand domain business logic and build reliable applied solutions from scratch.
- Work with LLM APIs, databases, internal systems, queues, logs, and access rights in a real production environment.
- Set up evaluation and observability: log call chains, measure result quality, configure alerts.
Requirements:
- Experience working daily with agents – Claude Code, Codex, Cursor, or a custom pipeline. As the primary workflow.
- Strong backend foundation: Python (asyncio), PostgreSQL/ClickHouse, queues, production systems under load.
- Practical experience with LLMs in production – tool calling, structured outputs, RAG, vector databases.
- Ability to determine where an agent works independently, where a human is needed, how to verify results, and how to put a system into operation.
- Good understanding of model limitations.
- English B2+, Russian for internal communication.
- Physically located outside the Russian Federation/Republic of Belarus, willingness to sign a B2B contract.
Conditions:
- Long-term B2B contract
- Rate up to $7000 net (depending on stack and experience)
- As the product grows – opportunity to delegate tasks and build your own team.
Contact: Anna Chernova, Telegram: Annachernova11