AI Engineer / LLM Engineer (AI-First | TypeScript)

Hello,

We are called People More because we treat our employees with respect, but also because the projects we work on are for people and should be easy and pleasant to use. We are technological, but we look at the bigger picture :)

The company is made up of people with a huge client base in the country and abroad, for whom we build projects from scratch (UX, UI, frontend, backend, mobile) or in part. We work directly for our clients and also support our partners in their own solutions. This ensures a wide range of projects and the ability to change! We work with clients all over the world.

For the project that we are working on with our partner, we are looking for AI Engineer / LLM Engineer (AI-First | TypeScript).

Your duties will include:

• Designing and developing production-grade agentic systems,
• Building workflows powered by LLMs and multimodal models,
• Implementing tool calling, routing, memory/context management, and structured outputs,
• Developing agent orchestration and multi-agent architectures,
• Evaluating output quality, debugging, and optimizing cost/performance,
• Supporting the development of an internal agent engine,
• Using agentic coding as a daily engineering workflow.

Requirements that must be met:

• Software engineering background,
• Experience building production systems leveraging LLM-based workflows beyond simple RAG or PoC implementations,
• Strong understanding of modern agentic patterns:
  • tool use,
  • planner-executor,
  • sub-agents,
  • single-agent vs multi-agent,
  • memory/context,
  • routing,
• Methodical approach to evaluations, tuning, and error tracking,
• Ability to track and optimize both cost and quality,
• Practical experience with tool calling and structured outputs,
• Practical MCP experience - both as a consumer and as a provider/server,
• Understanding of prompt injection and other risks related to agentic systems,
• Agentic coding as a daily workflow
• Very good English and Polish

Nice to have:

• Track record in model training, fine-tuning, or evaluations,
• Experience with self-hosted inference (e.g. vLLM),
• Public agentic/AI projects,
• Advanced RAG knowledge and understanding of related trade-offs.

Responsibilities

140 - 190 zł/h