Product Engineer

Department: Engineering

Location: Berlin

Employment Type: FullTime

About the Role

Join an early-stage AI company building tools for the global logistics industry. In this high-ownership role, you will work with the founding team across customer discovery, product development, and production systems, helping bring AI-powered workflows to users.

What You'll Do

• Speak with customers and translate their needs into product direction and technical architecture.

• Build full-stack features that combine user interfaces with LLM orchestration, agent workflows, and structured data pipelines.

• Develop reliable production AI systems, addressing latency, failure recovery, and evaluation.

• Turn early prototypes into robust systems, balancing speed with long-term quality.

• Contribute to technical vision and product strategy, and iterate on shipped work using user feedback.

What We're Looking For

• At least 3 years building and shipping full-stack web applications with Node.js and TypeScript in production, including recent hands-on React experience.

• Production experience building LLM-powered features, including prompt design, structured output parsing, context management, and error recovery.

• Experience with MongoDB schema design and query performance, distributed systems, and serverless infrastructure.

• Familiarity with vector stores, embedding pipelines, LLM orchestration, evaluation, and structured output validation.

• Experience using observability tools, including metrics and tracing, to debug production systems and asynchronous workflows.

• A record of owning customer-facing products from architecture through delivery and iteration, preferably in an early-stage environment.

• Strong customer communication skills; experience with complex B2B systems or logistics, supply chain, or operations is valuable.

Location

Hybrid, based in Berlin, Germany. The role is on-site Monday to Wednesday, with travel within Europe and an annual team offsite expected. Visa sponsorship is not available.