AI Architect

Your tasks

We are seeking an experienced AI Architect to lead the technical direction and delivery of reusable capabilities for Claude Code. In this role, you will bridge the gap between complex business requirements and cutting-edge AI execution by designing robust, production-grade skills and platform standards. Operating at the intersection of software engineering, agentic architecture, and process automation, you will act as a technical leader, guiding engineers and collaborating closely with business stakeholders to build scalable, non-deterministic AI solutions.

• Architecture & Technical Leadership
• Define Platform Standards: Establish and enforce quality bars, design patterns, shared infrastructure, and review criteria for well-formed AI skills.
• Architectural Oversight: Own the long-term technical roadmap, including evaluation frameworks, release strategies, integration patterns, and backwards compatibility.
• Team Guidance: Provide technical leadership and direction to engineering teams through code reviews, architectural unblocking, and mentorship.
• Strategic Decision-Making: Evaluate and justify platform-level choices regarding agent harnesses, integration architectures, and build-versus-adopt strategies.
• Skill Development & Engineering
• Design & Implementation: Build, test, and maintain production-ready, reusable Claude Code Skills, supporting scripts, validation logic, and context assemblies.
• Process Translation: Convert complex or ambiguous business domain workflows into structured, deterministic, and executable instructions for AI agents.
• Testing & Evaluation: Establish realistic test suites, evaluation scenarios, and acceptance criteria to measure and improve the reliability, accuracy, and usability of non-deterministic systems.
• Stakeholder Collaboration & Automation
• Requirements Discovery: Partner with business stakeholders and domain experts to analyze processes, constraints, and documentation to pinpoint automation opportunities.
• Workflow Optimization: Identify high-impact use cases across business operations that can be streamlined through custom AI agent workflows.

Requirements

Required Experience & Qualifications

• AI & Agentic Engineering: Deep understanding of modern AI patterns—including agentic workflows, Model Context Protocol (MCP), tool use, prompt and context engineering, and structured instruction design.
• Hands-on Agent Experience: Proven experience applying Claude Code (or comparable AI coding agent frameworks) to complex software engineering or business automation problems.
• Core Software Engineering: Solid background in building, reviewing, and maintaining production-quality software solutions using modern programming languages (e.g., Python, TypeScript).
• System Architecture & Ownership: Demonstrated ability to make and defend foundational platform decisions, design shared developer infrastructure, and govern platform-level standards.
• Evaluation of Non-Deterministic Systems: Hands-on experience testing, debugging, and improving non-deterministic LLM-driven applications.
• Business-to-Technical Translation: Ability to interface with non-technical stakeholders, extract operational knowledge, and translate it into robust technical designs.

Preferred Qualifications (Nice to Have)

• Claude Ecosystem Mastery: Direct experience authoring official Claude Code Skills, custom plugins, or Model Context Protocol (MCP) servers.
• LLM Evaluation Frameworks: Familiarity with evaluation and observability tools such as Promptfoo, LangSmith, or custom evaluation harnesses.
• Domain Knowledge: Exposure to enterprise business domains such as HR technology, compensation & benefits, payroll, or workforce analytics.
• Developer Tooling Background: Prior experience building internal developer platforms (IDP), CLI tools, or enterprise integration harnesses.
• Consulting & Public Artifacts: Background in technical consulting/professional services, or public contributions (open-source agent tooling, published skills/plugins, technical articles, or conference talks on agentic engineering).

Must have: AI, Claude, Python, TypeScript, Testing

Nice to have: LLM, CLI