About the project
Our client is the world's largest human resources consulting firm, headquartered in New York City, with operations across more than 130 countries and major branches in over 40 countries. The organization employs more than 25,000 people globally and provides services to the majority of Fortune 500 companies.
We are looking for an experienced Enterprise AI Architect to join a team responsible for designing and building reusable AI capabilities and Claude Code Skills that can be applied across enterprise business processes and software-development workflows.
The role combines AI engineering, software architecture, prompt and context engineering, and business process automation. The successful candidate will work closely with business stakeholders and domain experts to understand complex business processes and transform them into reliable, reusable and maintainable capabilities for AI agents.
This is a hands-on technical leadership position. In addition to developing AI capabilities, the Enterprise AI Architect will establish engineering standards, review technical solutions, guide other engineers, define evaluation and release strategies, and make platform-level architectural decisions.
The ideal candidate is a senior/principal-level AI engineer or architect who combines strong software engineering fundamentals with practical experience building AI-agent capabilities. This person should be comfortable moving between business discussions and low-level technical implementation, turning ambiguous business processes into robust AI workflows and establishing engineering standards for a broader AI platform.
This is not a purely strategic or conceptual architecture role. The successful candidate should be highly hands-on, capable of building and testing Claude Code Skills themselves while simultaneously providing technical leadership and defining the architecture, standards and engineering practices for the wider team.
Your responsibilities
- Work directly with business stakeholders and domain experts to understand business objectives, processes, constraints, requirements and expected outcomes.
- Analyse business requirements, process documentation, examples, templates, operational knowledge and other reference materials.
- Identify business workflows and use cases that can be automated, enhanced or standardized using Claude Code Skills and AI agents.
- Design, implement, test and maintain reusable Claude Code Skills and related AI capabilities.
- Create structured skill instructions, supporting scripts, reference materials, validation logic and reusable workflow components.
- Translate complex, ambiguous or highly contextual business processes into clear, structured and executable instructions for AI agents.
- Design AI workflows that are reliable, maintainable, reusable and suitable for enterprise-scale adoption.
- Evaluate AI skill behaviour using realistic business scenarios, test cases, acceptance criteria and repeatable evaluation frameworks.
- Analyse non-deterministic AI behaviour and continuously improve the accuracy, reliability, consistency and usability of AI-generated outputs.
- Establish standards and guidelines defining what constitutes a well-designed and production-ready AI Skill.
- Determine which capabilities should be implemented as individual skills and which should become shared infrastructure or reusable platform components.
- Define technical standards for skill development, testing, evaluation, versioning, release management and backward compatibility.
- Provide technical leadership to other engineers through architecture reviews, code and solution reviews, technical direction and problem-solving.
- Unblock engineers and support the team in resolving complex technical and architectural challenges.
- Make and justify platform-level technical decisions related to agent harnesses, integration architecture, reusable components and build-versus-adopt strategies.
- Own the overall technical shape of the AI capability and platform rather than focusing solely on individual skills.
- Establish and continuously improve the development lifecycle for AI capabilities, including design, implementation, evaluation, release and maintenance.
Our requirements
- Strong hands-on experience with Claude Code in real-world software-development or business-workflow scenarios.
- Proven practical experience creating Claude Code Skills or comparable reusable capabilities for AI coding agents.
- Strong understanding of modern AI engineering and agentic AI patterns, including: agentic workflows, AI agent orchestration, tool use, prompt engineering, context engineering, structured instruction design, reusable AI capabilities.
- Strong professional software engineering background with the ability to design, implement, review, test and maintain production-quality solutions.
- Experience translating business requirements and domain knowledge into technical designs, structured workflows and executable AI instructions.
- Practical experience evaluating, troubleshooting and improving non-deterministic AI/LLM-based solutions.
- Strong understanding of software architecture and the ability to make decisions affecting an entire AI platform or capability.
- Experience defining engineering standards, reusable components and architectural patterns.
- Ability to establish clear quality criteria for AI Skills and determine what should be accepted, redesigned or rejected during technical review.
- Ability to distinguish between functionality that belongs within an individual skill and functionality that should be implemented as shared infrastructure.
- Strong technical leadership skills, including experience reviewing other engineers' solutions, providing technical direction and resolving blockers.
- Hands-on experience with Claude Code plugins and/or MCP servers.
- Experience with LLM evaluation and testing frameworks such as Promptfoo, LangSmith or custom evaluation harnesses.
- Experience working in consulting, professional services or client-facing technology environments.
- Experience building internal developer tooling or establishing a new technical platform/capability from the ground up.
- Public technical artefacts, such as: open-source AI/agent tooling, published Claude Code Skills or plugins, MCP servers, conference presentations, technical articles or blog posts focused on agentic engineering.