Backend Python Developer
About the platform:
• Modular Intelligent Data Automation (MIDA) is a platform developed by Deloitte Tax & Legal Germany that turns incoming data into decision-ready data for downstream processes and systems.
• Modular processing chain per use case, with steps such as classification, extraction, validation, enrichment and analysis, each running on deterministic scripts, machine learning or LLM skills.
• Production-ready availability of Anthropic and OpenAI LLMs for skill-based processing via Azure AI Foundry and Gemini Agent Platform compliant with Deloitte standards.
• One of the first platforms within Deloitte to run Claude-based skills compliantly in production, with LLM skills delivering structured data directly within the processing chain.
• Multiple use cases in production since 07/2024, currently processing approx. 250,000 datasets per year.
• Python back-end with REST APIs and RabbitMQ, Pydantic and Pydantic AI, Angular frontend, Azure Kubernetes Service, Azure Cosmos DB and Blob Storage, and AI services from Azure and Google Cloud.
Role / Responsibilities:
• Back-end support for a full-stack team, working closely with the technical team lead.
• Conceptualization and delivery of new features, from technical concept to production-grade Python code.
• Design and delivery of LLM skills and their integration into the processing chain.
• Python back-end pull request reviews and quality assurance, with actionable feedback and joint merge decisions with the technical team lead.
• Support of architecture and technical design decisions, plus mentoring of other developers.
• Contribution to engineering practices and development processes.
Required Skills
Required Languages
🇬🇧 English