Enterprise AI Architect

Exadel Inc (Website)•Town of Poland, NY

1d•Hybrid

About The Position

This role is for an Enterprise AI Architect at Exadel, a global tech company with extensive experience in AI platforms and digital transformation. The client is the world's largest human resources consulting firm, serving 97% of Fortune 500 companies and operating in over 130 countries.

Requirements

Strong hands-on experience using Claude Code in real software-development or business-workflow scenarios.
Practical experience creating Claude Code Skills or comparable reusable capabilities for AI coding agents.
Strong understanding of modern AI engineering patterns, including agentic workflows, tool use, prompt engineering, context engineering, and structured instruction design.
Proven hands-on software engineering experience and the ability to build, review, test, and maintain production-quality solutions.
Ability to transform business requirements and domain knowledge into clear technical designs and executable workflows.
Experience evaluating and improving non-deterministic AI solutions.
English level Upper-Intermediate.

Nice To Haves

Hands-on experience with Claude Code, and specifically with authoring Claude Code Skills, plugins, or MCP servers.
Evaluation tooling for LLM systems (promptfoo, LangSmith, custom harnesses).
Consulting or professional services background.
Domain exposure to HR technology, compensation & benefits, payroll, or workforce analytics.
Experience building internal developer tooling or standing up a platform capability.
Public artefacts: open-source agent tooling, published skills or plugins, conference talks, technical writing on agentic engineering.

Responsibilities

Collaborate directly with business stakeholders and domain experts to understand their objectives, processes, constraints, and expected outcomes.
Analyze business requirements, documentation, examples, templates, and other reference materials provided by stakeholders.
Identify workflows and use cases that can be implemented or improved through Claude Code Skills.
Design, implement, test, and maintain reusable Claude Code Skills.
Create clear skill instructions, supporting scripts, reference materials, validation logic, and reusable workflow components.
Translate complex or ambiguous business processes into structured, executable instructions for AI agents.
Evaluate skill behavior using realistic scenarios, test cases, and acceptance criteria.
Improve the reliability, accuracy, usability, and maintainability of developed skills through iterative testing and stakeholder feedback.
Own the technical shape of the whole capability, not individual skills: standards, shared components, evaluation and release strategy, backwards compatibility.
Define and enforce what a well-formed skill looks like, what belongs in shared infrastructure, and what gets rejected in review.
Make and justify platform-level decisions - including choices around agent harnesses, integration architecture, and build-versus-adopt strategies, while providing clear technical rationale for foundational decisions that impact the entire platform.
Provide technical leadership over other engineers: review, direction, unblocking.