Senior AI Engineer

In a partnership with a Berlin-based startup that has successfully established a leading digital trading platform for industrial metals across Europe, we are looking for a heavy-hitting Senior AI Engineer to join our team. You aren’t just someone who calls APIs; you are an architect of intelligent systems who understands the "why" behind the "how."

You will be responsible for building, fine-tuning, and deploying sophisticated AI agents and custom models within the GCP ecosystem. If you have a deep obsession with agentic workflows, the Model Context Protocol (MCP), and the nuances of model evaluation, we want to talk to you.

Key Responsibilities

• Architect Agentic Systems: Design and implement complex agentic layers and workflows using LangChain and LangGraph to solve multi-step reasoning problems.
• Model Engineering: Select, deploy, and fine-tune open-source models (Llama, Mistral, etc.) alongside proprietary LLMs (Gemini, OpenAI).
• R&D & MCP: Leverage the Model Context Protocol (MCP) to integrate local and remote resources into agentic environments.
• Production Excellence: Implement rigorous model evaluation frameworks, regression testing, and observability using LangSmith.
• System Design: Build scalable AI infrastructure on GCP, ensuring seamless integration between custom models and cloud-native services.

Technical Requirements

• Experience: 3–5+ years of high-level experience in AI/ML engineering (Very Senior level).
• GenAI Stack: Expert-level command of Gemini, OpenAI, and Llama.
• Orchestration: Deep expertise in LangChain and LangSmith for building and monitoring agentic tools.
• Cloud: Professional experience with GCP (Vertex AI, Cloud Run, GKE).
• Coding: Python mastery is a given.
• Fine-Tuning: Proven track record of leveraging and fine-tuning open-source models for specific domain tasks.
• Testing: Strong focus on model evaluation, bench-marking, and regression testing.
• Nice to Have: Deep learning frameworks like PyTorch or TensorFlow and a solid grasp of classic ML concepts.

Interview Process

1. Technical Deep-Dive: A 1-hour session with our Engineering Manager focusing on system design, agentic logic, and live problem-solving.
2. Culture & Vision: A conversation with our Founder or HR lead to discuss soft skills, values, and long-term fit.