- Opened · 30d
- +0
- Closed · 30d
- −0
- All ml / ai
- 6,674open
About us
We're an independent European digital product studio with a strong track record of delivering market-defining strategies and products for enterprises and governments. With about 200 talented people across three countries, premium AI capabilities and a diverse portfolio of international clients, we define, design and ship high-quality digital products, strategies and software. From global cloud-based platforms and large-scale apps to medical devices, banking software and augmented reality experiences. In a nutshell: if it lives and breathes digital, it’s what we do best.
Become a part of our team
At In The Pocket, you join a multidisciplinary team of designers, engineers, strategists and data specialists who care about the work as much as the people they work with. We work in small, autonomous teams that stay close to our clients and their users, and we give everyone the space to take ownership, speak up and grow. You’ll learn from colleagues across our studios, ship products that reach thousands or even millions of people and help shape what digital can do for leading organisations in Europe.
What you’ll do
We are looking for an AI Engineer to join In The Pocket. Foundation models have made a lot of things possible that weren't possible a few years ago. Your mission as an AI Engineer is to harness that power and turn it into a capable, reliable and trustworthy product that makes people happy and businesses grow.
- Design and build AI-powered features on top of foundation models: agents, multimodal retrieval systems, assistants, document analysis, advanced search capabilities, and more.
- Guide AI solutions from early experiments and prototypes to production deployment, integrating them into existing applications, workflows, and enterprise systems.
- Design and implement agentic workflows, tool integrations, and context management systems that allow AI to interact reliably with the world around it.
- Build evaluation into your work from the start. Define success criteria, measure performance, investigate failures, and continuously improve the quality of your solutions.
- Make AI systems production-ready, with attention to reliability, observability, latency, scalability, security, and cost.
- Make thoughtful engineering decisions about orchestration, model selection, context management, and human oversight. Know when AI adds value and when conventional software is the better choice.
- Work directly with clients and end-users to define evaluation criteria, support user testing, capture feedback, and translate insights into improvements that make AI solutions more effective and reliable.
- Share what you learn in our AI engineering practice, whether that's a demo, an internal session or a written-up experiment, and build on what your colleagues have learned.
Who you are
- You’re a passionate engineer that wants to create elegant solutions for complex problems. You either have a Master's in Computer Science, AI or a related field, or you have a proven track record as an AI engineer.
- You have a deep knowledge of and hands-on experience with designing, creating and operating LLM-based solutions. You understand the capabilities, limitations and risks of using large language models, and you know how to build a reliable product on top of them (prompt engineering, context engineering, harness engineering,…).
- You have a deep understanding of concepts such as agents and tools, and you have experience with important industry standards such as MCP, agentskills, A2A,…
- You have experience with retrieval and reranking, embeddings, chunking strategies and vector databases.
- You’re well aware of common software engineering best practices (CI/CD, automated testing, infrastructure as code,…), and you apply them naturally.
- You know what secure software looks like and you apply the same standard to AI-generated code as to your own. Familiar with OWASP, and aware of the risks that are specific to LLM systems (prompt injection, data leaking through retrieval, permissions that are too broad)