You are interested in the position as AI Engineer (Agentic Systems & Infrastructure) at PDR.cloud GmbH?
On the job platform Instaffo, you can apply for this and other jobs with ease. Create your profile in under 3 minutes and start the application process with just a few clicks.
Please note that the position is only available to applicants from Germany.
The language requirements for this position are: German - Fluent, English - Fluent.

In this role, you will design and build an agentic AI system to power automated customer support. By combining conversational AI, multi-agent orchestration, and scalable infrastructure, you’ll create solutions that deliver faster, smarter answers while reducing manual effort. You’ll work at the forefront of GenAI, transforming cutting-edge research into practical tools that improve efficiency and customer satisfaction. Aligned with our mission to empower customers to handle high volumes of requests efficiently, you’ll enable them to stay focused on the work that truly matters.

Activities

About the Role

As an AI Engineer (Agentic Systems & Infrastructure), you will play a central role in designing and implementing our next generation of AI-powered customer support solutions. Your mission will be to build an agentic system that acts as an intelligent tool to automate customer requests, enabling our customers to manage high workloads while staying focused on the work that truly matters.

This position combines AI research, backend engineering, and system design. You will architect and develop scalable, production-ready systems that bring together conversational AI, multi-agent frameworks, and enterprise-grade integrations. These systems will not only provide immediate value by automating customer support interactions but will also serve as the foundation for future AI-driven features across our products.

Key Responsibilities

• System Architecture & Design: Define, plan, and implement end-to-end agentic systems that orchestrate large language models (LLMs), conversational flows, and supporting services.
• Conversational AI & Multi-Agent Orchestration: Build and optimize AI agents capable of managing complex dialogues, retrieving knowledge, and coordinating across multiple tools or services.
• Integration & APIs: Develop and maintain robust integrations with internal and external systems via REST and gRPC, ensuring interoperability and extensibility.
• MLOps & LLMOps: Implement best practices for deployment pipelines, monitoring, logging, autoscaling, and security to ensure that AI solutions remain stable and cost-efficient in production.

Requirements

Your Profile / Requirements

• Proven experience in building conversational AI and/or multi-agent systems (e.g. with LangChain, AutoGen, CrewAI, or similar frameworks).
• Strong programming skills in Python (TypeScript/Node.js is a plus).
• Solid understanding of REST and gRPC APIs, including design, integration, and scaling.
• Experience deploying and serving LLMs with frameworks such as vLLM, Triton Inference Server, or Hugging Face TGI.
• Familiarity with cloud platforms (AWS, Azure, or GCP) and containerization (Docker, Kubernetes).
• Knowledge of MLOps/LLMOps practices, including CI/CD pipelines, monitoring, logging, and autoscaling.
• Understanding of vector databases (e.g. Pinecone, Weaviate, Milvus) and RAG architectures.
• Strong grasp of system design and ability to translate business needs into scalable technical solutions.
• Commitment to security, reliability, and performance in production AI systems.
• Excellent problem-solving skills and the ability to work in a cross-functional, fast-paced environment.

Nice to Have

• Experience with GPU optimization (TensorRT-LLM, CUDA) or distributed systems.
• Contributions to open-source AI frameworks.
• Background in customer support automation, enterprise integrations, or SaaS platforms.

Team

You will become part of our dedicated, interdisciplinary development team. We place great value on open communication, mutual support, and a constructive feedback culture. While you can work primarily remotely, we meet regularly in Berlin-Schöneberg to develop ideas, align strategies, and, of course, enjoy lunch together. Our team embraces a flat hierarchy, short decision-making paths, and a friendly, collaborative atmosphere that leaves room for creativity and personal growth.

Application Process

Our application process is transparent, lean, and personal:

• Initial Call (Remote) – A short conversation to get to know you and your motivation.
• Technical Interview – A technical discussion with our engineers, possibly including a code review or a small practical task.
• Team Interview – An exchange with future colleagues, giving you the chance to ask questions and get to know our working environment.
• Final Interview & Offer – Alignment on details and the earliest possible starting date.

We value fast feedback and will guide you through the entire process with openness and respect.

About the Company

PDR.cloud wurde 2018 in Berlin gegründet.

Wir bieten KFZ-Reparaturdienstleistern eine cloud-basierte Software zur Abrechnung und Steuerung von Schäden.

PDR.cloud vereint die bekannten Funktionen eines klassischen Dealer-Management-Systems mit smarten Lösungen und zeitgemäßer IT-Architektur.

Unsere Mission ist es für unsere Kunden eine vollständig digitale Werkstatt zu realisieren und den Arbeitsalltag unserer Anwender trotz immer komplexer werdenden Schaden-Prozessen zu vereinfachen.