Practice Architect AI/ML

Solution architecture

The Practice Architect, Level 1 is responsible for translating enterprise business problems into governed, production-ready AI solutions. They design and build agentic systems on the Claude platform, focusing on multi-agent orchestration, Model Context Protocol (MCP) integrations, and retrieval pipelines. They also partner with client stakeholders to ensure every solution is secure, scalable, and compliant.

Agentic system development

• Build and productionize agentic workflows using Claude, MCP servers, and agentic frameworks (LangChain, CrewAI, AutoGen, Semantic Kernel, or Strands).
• Lead adoption of AI coding tools such as Claude Code and Cursor AI, establishing patterns and guardrails that raise engineering velocity and quality.
• Architect and deploy solutions across GCP, AWS, or Azure, applying well-architected principles for reliability, cost, and security.

Cloud architecture

• Architect and deploy solutions across GCP, AWS, or Azure, applying well-architected principles for reliability, cost, and security.

AI FinOps & cost governance

• Establish cost visibility and spend governance for AI workloads across multi-cloud estates (GCP, Azure, WCNP).
• Model token economics and build guardrails that keep agentic systems economically sustainable at scale.

Security & guardrails

• Design security into every solution: secure-by-default architectures, least-privilege agent and tool permissions, input/output mediation, and defenses against prompt injection and excessive agency, anchored to the OWASP LLM Top 10.

Governance & compliance

• Embed responsible-AI controls aligned to NIST AI RMF, ISO/IEC 42001, and the EU AI Act.
• Define human-in-the-loop review gates, audit logging, and output provenance so systems are defensible to auditors and clients.

Reusable assets

• Contribute to the shared CoE library of skills, MCP servers, and reference architectures so patterns are reused across the client portfolio.

Client partnership

• Engage directly with client architects and stakeholders;
• Present designs, run technical workshops, and mentor engineers on the account.

Required Qualifications

• 5+ years of professional software engineering experience, with a strong foundation in Python and modern software design practices.
• 3+ years working in applied AI, including hands-on experience with LLM platforms (Claude strongly preferred) and AI development tools such as Cursor AI or Claude Code.
• Demonstrated experience building AI applications: prompt engineering, RAG pipelines, agent orchestration, or MCP / tool-calling integrations.
• Cloud architecture experience on at least one major platform (GCP, AWS, or Azure), including compute, storage, networking, and IAM fundamentals.
• Proven ability to design and document solution architectures and communicate technical trade-offs to both engineers and business stakeholders.
• Working knowledge of application and AI security: secure design principles, least privilege, and guardrails against common LLM risks (e.g., prompt injection, insecure output handling).
• Bachelor’s degree in Computer Science, Engineering, or a related field, or equivalent practical experience.

Preferred Qualifications

• Cloud architect certification on GCP, AWS, or Azure; multi-cloud exposure is a strong plus.
• Experience deploying agentic or LLM systems into regulated environments (financial services, healthcare, or retail).
• Experience with FinOps practices or AI cost governance: token economics, multi-cloud spend optimization, or model/tooling cost management.
• Familiarity with AI governance frameworks: NIST AI RMF, ISO/IEC 42001, or the EU AI Act.
• Security background: secure code review, threat modeling, or OWASP LLM Top 10 assessment of AI workloads.
• Hands-on experience building MCP servers or custom agent tools.
• Contributions to internal enablement: reusable skills, reference architectures, or engineering playbooks.