KERING Principal AI Engineer-Agentic Systems

How you will contribute

AI Architecture & Business Understanding

•       Work with product managers and business experts to understand workflows, assess where agents can help, and define expected outcomes, operating limits and human decision points.

•       Design agent systems on shared orchestration and knowledge services, covering model selection, prompts, retrieval, state, memory and tools. Balance task quality, reliability, response time and cost.

•       Write, review and debug production code in Python and other languages as needed. Resolve failures across agent logic, business APIs and services while preserving API contracts, authorization and traceability.

Evaluation & Continuous Improvement

•       Build evaluation suites from business scenarios, production failures and user feedback. Work with domain experts and QA to assess task completion, response quality, permission boundaries and the correctness of actions.

•       Use execution traces and evaluation results to diagnose weaknesses in prompts, context, tools and agent logic. Compare changes against a baseline and check for regressions before release.

•       Develop repeatable optimization workflows with versioned prompts, configurations, code and tests. Use reviewed releases and recovery plans to introduce improvements to production agents.

Technical Leadership & Partner Collaboration

•       Review delivery partners’ architectures, code, prompts, workflow configurations and evaluation evidence. Resolve design and quality issues with them and support technical acceptance within Kering’s engineering standards.

•       Lead the technical transition of selected partner-built agents, retaining source code, workflow definitions, evaluation assets and operating knowledge so the team can reproduce, support and improve them internally.

•       Mentor engineers and establish reusable agent patterns with the development lead, platform engineers and QA. Guide the use of shared model access, knowledge services and approved business APIs.

Who you are?

Experience & Technical Skills

•       Bachelor’s degree or above in Computer Science, Software Engineering, or a related field.

•       Extensive software or AI engineering experience, typically eight or more years, including hands-on technical leadership of AI agents used in production.

•       Strong Python and software engineering fundamentals, with a track record of solving problems across programming languages and technology stacks. Able to understand, modify and debug unfamiliar codebases.

•       Practical experience using AI-assisted development to investigate unfamiliar codebases, implement changes and build tests, with independent judgment in reviewing the results for correctness, security and maintainability.

•       Deep practical understanding of tool-using agents, context management, retrieval-augmented generation (RAG), state, memory and orchestration; able to choose between configured workflows and custom code.

•       Experience designing agent evaluations, investigating production traces, and combining automated checks with expert review to assess quality and guide improvements.

•       Experience supporting production applications: containerized deployment, CI/CD, identity and access control, monitoring, incident diagnosis and recovery.

•       Ability to design and assess business API contracts (REST/OpenAPI), including user and tenant permissions, input validation, action confirmation and safe handling of retries.

•       Evidence of translating business workflows into technical decisions and leading design discussions with engineers, product teams and delivery partners.

•       Fluency in Mandarin and good working proficiency in English.

Preferred Experience

•       Experience taking over externally developed AI applications and maintaining them through subsequent releases.

•       Experience with Dify or comparable orchestration platforms, Model Context Protocol (MCP), or automated prompt and agent optimization validated through regression testing.

Personal Attributes

•       You stay close to implementation and follow difficult problems through to resolution.

•       You explain technical trade-offs clearly and build agreement across teams through evidence and practical examples.

•       You seek feedback from users and engineers, and turn what you learn into better systems and shared working practices.