Python Developer - Project Delivery Specialist

Deloitte Deloitte · Consulting

Develops and orchestrates LLM workflows, RAG pipelines, and multi-agent systems using Python, LangChain, and LangGraph, focusing on production deployment and operational ownership.

What you'd actually do

  1. Build and orchestrate LLM workflows in Python using LangChain/LangGraph, including tool/function calling, structured outputs, error handling, and performance/cost tuning.
  2. Design and operate RAG pipelines: data ingestion/chunking, embeddings + vector search, retrieval/reranking, grounding/citation, and evaluation to reduce hallucinations.
  3. Implement multi-agent and memory systems: agent roles/handoffs, state management, short/long-term memory persistence, conversational checkpoints, and guardrails/observability in production.
  4. Documentation & Knowledge Sharing: Create and maintain comprehensive documentation for cloud architectures, configurations, and operational procedures. Mentor junior engineers and contribute to knowledge sharing within the team. Communicate regularly with Engagement Managers (Directors), project team members, and representatives from various functional and / or technical teams, including escalating any matters that require additional attention and consideration from engagement management

Skills

Required

  • Python
  • LangChain
  • LangGraph
  • RAG
  • Multi-agent systems
  • Memory systems

Nice to have

  • Public cloud provider certifications (e.g., AWS Certified Solutions Architect, AWS Certified DevOps Engineer, Google Cloud Professional Cloud Architect, Google Cloud Professional DevOps Engineer).
  • Serverless computing (e.g., AWS Lambda, Google Cloud Functions).
  • Configuration management tools like Ansible.
  • Database services (relational and NoSQL) in the cloud.
  • Disaster recovery and business continuity strategies in a multi-cloud environment.
  • GitOps principles and tools.
  • Analytical ability to manage multiple projects and prioritize tasks into manageable work products
  • Can operate independently or with minimum supervision
  • Excellent Written and Communication Skills
  • Ability to deliver technical demonstrations

What the JD emphasized

  • 6+yrs Python (strong proficiency)
  • 6+yrs LangChain & LangGraph
  • 6+yrs RAG (Retrieval-Augmented Generation)
  • Multi-agent systems
  • Memory systems (long-term/short-term, persistence, conversational checkpoints)

Other signals

  • LLM workflows
  • RAG pipelines
  • Multi-agent systems