Full-Stack Engineer (Python)

Build the platform where teams design and run AI-powered workflows over documents and data. Join the Global Development AI engineering team. At Deloitte.

• build features end-to-end across an async Python backend and a React/TypeScript frontend;
• orchestrate multi-step, LLM-powered pipelines that process documents and structured data;
• design the data models, schemas and safe migrations that keep workflow execution reliable at scale;
• create interactive, node-based canvas UIs where users visually design and monitor their workflows;
• apply LLM application patterns such as prompting, structured outputs and multi-step reasoning;

own the full path of a feature, from data model to API to on-screen experience.

Naturally   you like to experiment  and you continue until you have found the solution .But most importantly, be the true you. The one and only you. With your personal strengths, view of the world and unique personality. You need the following qualifications for the position of Full-Stack Engineer:

• university degree in computer science, software engineering, or a related background;
• excellent command of English;
• solid experience building full-stack applications with Python and React/TypeScript, and curiosity about how LLMs behave in real products;
• comfort modelling backend data and execution flow, including schema design and safe migrations;
• you care about reliability, because a workflow that fails silently is worse than one that fails loudly;
• you can turn a fuzzy requirement into a working feature without waiting to be told every step.

As a Full-Stack Engineer on the Global Development AI team, you shape how teams across Deloitte put AI to work. You build the blocks, snap them together into pipelines, and watch other teams create things you did not expect. Those pipelines power real work: automated due diligence, tabular data review and report generation. The platform turns complex, multi-step reasoning into something people can design visually and trust to run reliably. You will not sit in a single layer. You own features from the data model up to the screen, so the choices you make are felt directly by the people who use the platform every day.