Senior AI Engineer - Supply Chain

Cognite Cognite · Industrial · India · Center of Excellence

Senior AI Engineer focused on building and enhancing packaged, customer-ready agentic workflows for supply chain automation. The role involves developing multi-agent systems that leverage tool-use, memory, and reasoning to interact with a knowledge graph, automate complex decisions, and provide operational insights and recommendations. Requires strong full-stack engineering skills, experience with LLM-driven agent frameworks, and understanding of industrial and supply chain data.

What you'd actually do

  1. Develop modular, multi-agent systems using ATLAS AI that leverage tool-use, skills, memory, and multi-step reasoning (Chain-of-Thought) to interact with the CDF Knowledge Graph for solving supply chain problems. This should be product ready and handed to the product and delivery organisation.
  2. Design and build precise, reliable tools that can query and traverse the supply chain knowledge graph — handling complex relationships, multi-hop lookups, and returning accurate structured data. The supply chain agents should be able to reason over it.
  3. Craft advanced prompt chains and evaluation frameworks to ensure agents provide accurate, operationally safe recommendations for industrial use cases.
  4. Operates with high autonomy and minimal guidance to build functional prototypes directly on customer data or synthetic data, test with customers and SMEs and support integration with customers working along with a value-delivery team.

Skills

Required

  • 5-8 Years in AI/ML Engineering
  • Proven track record of delivering production-grade AI features
  • Hands-on experience with LLMs, embeddings, prompt engineering, and related techniques
  • Experience building LLM-driven or agentic systems (e.g., using LangChain or similar frameworks)
  • Ability to query and navigate complex relational and graph data models
  • You can work across the stack and own delivery of functional, user-facing solutions: Frontend: React (or similar) + TypeScript, Backend: Python, Kotlin, or similar for orchestration and API integration
  • Deep understanding of how to work with "imperfect" industrial data and cross functional supply chain data, including time series and sparse metadata
  • A "show, don't tell" approach
  • Bias towards rapid iteration
  • You can take a schema (like the Integrated Supply Chain model) and independently identify how to build an agent that provides value from it
  • Experience using coding assistants (Cursor, Claude Code, Copilot, or similar) and comfortable using LLMs to accelerate development, debugging, and documentation tasks
  • You communicate clearly

What the JD emphasized

  • packaged , customer ready solution
  • product ready
  • customer ready

Other signals

  • building agents capable of traversing the Cognite Data Fusion (CDF) Knowledge Graph to automate complex supply chain decisions
  • packaged , customer ready solution of agentic workflows
  • reason across these domains