Senior Full-stack Engineer (3d)

Autodesk Autodesk · Enterprise · London, United Kingdom +2

Senior Full-Stack Engineer to build AI-native applications with agentic workflows. The role involves designing and implementing backend logic for AI agents to interact with external systems, building human-in-the-loop interfaces for user collaboration, and optimizing the entire stack for performance, especially LLM streaming latency.

What you'd actually do

  1. Lead the development of full-stack features, from database schema design to high-fidelity UI implementation
  2. Design and implement backend logic that allows AI agents to interact with APIs, databases, and long-running processes autonomously
  3. Build "Human-in-the-loop" interfaces that allow users to monitor, steer, and collaborate with AI agents in real-time
  4. Architect secure and performant backend systems using best practices and design patterns, to support high-concurrency AI operations
  5. Manage complex, asynchronous application states where data may be updated concurrently by both users and background AI agents

Skills

Required

  • React
  • TypeScript
  • modern CSS/HTML
  • RDBs (e.g. PostgreSQL)
  • Object Stores (e.g. s3)
  • RESTful and/or GraphQL APIs
  • modern agentic development environments (e.g., Cursor, Claude Code, Windิง)
  • distributed systems
  • serverless architectures
  • relational database modeling
  • CI/CD practices and tools (e.g. Jenkins)
  • cloud providers for compute/network (e.g. AWS ALBs/ECS)
  • technical LLM capabilities and practical, user-facing product features

Nice to have

  • agentic orchestration frameworks (e.g., LangChain, LangGraph, CrewAI, or PydanticAI)
  • Agentic UX patterns, such as ReAct, streaming UI updates, generative UI, and proactive notifications
  • monitoring and evaluation tools for AI agents (e.g., LangSmith, Braintrust, or Helicone)
  • Tools/Functions for LLMs to interact with third-party SaaS platforms or internal databases
  • durable runtime frameworks such as temporal or similar
  • Supabase e.g. Row Level Security (RLS), and database migrations
  • pgvector or other vector stores for RAG (Retrieval-Augmented Generation) workflows
  • WebSockets or Supabase Realtime

What the JD emphasized

  • AI-native applications
  • agentic workflows
  • AI agents
  • interact with external tools
  • Human-in-the-loop interfaces
  • AI agents
  • high-concurrency AI operations
  • LLM streaming

Other signals

  • AI-native applications
  • agentic workflows
  • AI agents proactively execute tasks
  • interact with external tools
  • solve complex user problems
  • Human-in-the-loop interfaces
  • monitor, steer, and collaborate with AI agents
  • high-concurrency AI operations
  • minimize latency in LLM streaming