Junior Python Developer, Agentic AI

The company you would be joining Reply is a network of highly specialised companies operating in digital transformation, cloud, data and artificial intelligence. In 2025 the group closed at EUR 2,483.6 million in revenue with 16,624 people, up from EUR 277.2 million and 2,272 people in 2007. At the end of H1 2026 the group counts 17,297 people. Growth has been continuous, organic and profitable for close to two decades. The model is deliberate. Instead of one large generalist consultancy, Reply is built as a federation of focused companies, each one specialised in a technology, a platform or an industry, working together on large multi-country programmes. The practical consequence for an engineer is that you are surrounded by people who go deep rather than wide, and that the project you work on is staffed by the specific companies that own that competence, whether they sit in Brussels, Luxembourg, Milan, London or Düsseldorf. Reply holds the highest partnership tiers with the major technology vendors, including AWS Premier Partner, Google Cloud Premier Partner, Microsoft Global Partner, Oracle Global Partner, Salesforce Summit Partner, SAP Gold Partner and Adobe Platinum Partner, and works directly with the leading players in AI including Anthropic, OpenAI, Mistral, NVIDIA and Databricks. This gives our engineers early access to platforms, roadmaps and product teams, often before those technologies reach the wider market. The team You would join the Belgian AI transformation team, a small, senior group delivering agentic AI systems for major financial institutions and large corporates in Belgium and Luxembourg. The work is production work: multi-agent architectures, retrieval and document intelligence pipelines, tool-calling agents integrated into core business processes, and the engineering discipline required to run all of that in regulated environments. The team operates as one of the group's innovation teams. Technology choices move quickly, the client base is demanding, and the standard for what gets shipped is high. What you will do Design and build backend services and agentic components in Python, from prototype through to productionIntegrate LLMs and agent frameworks with enterprise systems through APIs, message queues and data storesWork with SQL and NoSQL databases, vector stores, containerised deployments and cloud services (AWS, Azure, GCP depending on the client. or on-prem)Contribute to architecture and design discussions, and to the evaluation of new models, frameworks and toolsWork directly with clients alongside senior engineers and architects, and with colleagues from other Reply companies on multi-country programmes What we are looking for A degree in Computer Science, Computer Engineering, Software Engineering or a comparable technical disciplineSolid Python fundamentals, demonstrated through academic projects, internships, personal projects or open source contributionsWorking knowledge of the technologies that support modern development: relational and NoSQL databases, REST APIs, message queues, Docker or equivalent container tooling, GitEnglish at professional level, plus French or Dutch at professional level. Client work in Belgium and Luxembourg requires itGenuine curiosity and the will to learn fast. The ability to work in a team of specialists that wants to make you growExperience with LLM frameworks, cloud platforms or agentic architectures is welcome but not required. What is required is the aptitude and the drive to reach that level quickly, with support. What we offer Continuous training, certification paths on the platforms of our technology partners, and access to Reply's internal knowledge network across all group companiesInternational exposure: Reply operates in 15 countries and project teams are frequently cross-border, with opportunities to work with and travel to other Reply companiesA career path built on technical depth.Work on some of the most advanced AI programmes currently running in the Belgian and Luxembourg financial sector How to apply Apply through this posting with your CV. Links to GitHub, personal projects or technical writing are read and they count.