• Lead AI engagements end-to-end: discovery, architecture, delivery oversight, executive updates, and knowledge transfer — you are the technical owner from first workshop to final handoff.
• Serve as the primary technical point of contact for client stakeholders, including VPs, CIOs, and CTOs; run workshops, steering committees, and difficult conversations with credibility and calm.
• Architect agentic systems, RAG pipelines, and AI-augmented workflows across the Microsoft, OpenAI, and Anthropic ecosystems — Azure OpenAI, Azure AI Foundry, AI Search, Copilot Studio, the OpenAI platform (Responses API, Agents SDK), and the Anthropic platform (Claude API, Claude Agent SDK, MCP, skills) — choosing and composing them based on client fit, not vendor loyalty.
• Build and maintain the reference architectures, skills, prompts, and MCP servers the team reuses across clients.
• Help shape our growing OpenAI and Anthropic practices alongside our Microsoft work — enablement content, reference implementations, and the technical narrative we take to clients and partners.
• Pair with delivery leads and engineers on the hard parts — review designs, debug production issues, sharpen evals and guardrails.
• Translate between executive strategy and engineering reality; write clearly, present crisply, and disagree well.
• Contribute to internal IP: opinionated docs, short demo videos, and enablement sessions that upskill the rest of the team on what is actually working.
• Help shape hiring, tooling, and delivery practices as the AI landscape continues to shift.

• 8+ years of enterprise software, data, or AI architecture experience, including at least 2 years architecting and shipping generative AI or agentic systems to production.
• Demonstrated leadership of complex, multi-stakeholder AI or software initiatives as the senior technical owner. This can come from either track — weight them equally: consulting or professional services (scoping, SOWs, delivery team coordination, client escalations), or an industry role leading internal programs end-to-end with business stakeholders, steering committees, and cross-functional delivery teams.
• Excellent stakeholder communication — written and verbal, to technical and executive audiences, whether those stakeholders are external clients or internal business partners. You can facilitate a whiteboard session, present a design review to a skeptical CIO (or your own VP of Engineering), and write a clean architecture document in the same week.
• Hands-on proficiency with the Azure AI stack: Azure OpenAI, Azure AI Foundry / AI Studio, AI Search, Document Intelligence, and at least one orchestration framework (Semantic Kernel, LangGraph, or equivalent).
• Hands-on proficiency with the OpenAI platform: Responses API, Agents SDK, function calling, structured outputs, and evals — you have shipped something on it, not just read the docs.
• Hands-on proficiency with the Anthropic platform: Claude API, tool use, Claude Agent SDK, MCP, and skills — again, demonstrated through work you can walk us through.
• Informed judgment on when to reach for which model and platform — GPT family, Claude family, Gemini, open-weights — and the trade-offs (capability, cost, latency, data residency, eval posture) that drive the call.
• Microsoft Certified — Azure AI Engineer Associate and Azure Solutions Architect Expert (or willingness to earn within 90 days of hire).
• Strong software engineering fundamentals — you can still read, write, and review code. You are not an architect who stopped building.
• Able to work remotely within the United States with reliable overlap for Eastern and Central business hours.

• A visible track record of building agentic solutions — production or substantive prototypes, not just demos. Bring repos, architecture walkthroughs, recorded demos, or case studies. We want to see what you have shipped, ideally across more than one model family.
• AI coding tools are a core part of your daily work — Claude Code, Cursor, GitHub Copilot, Codex, or equivalent. Bring receipts: agents you have configured, skills or prompts or slash commands you have authored, and commit history that reflects how you actually work with them.
• You have built agentic systems with real tool use — MCP servers, skills, sub-agents, multi-step planning and reflection loops, or comparable patterns on OpenAI Agents SDK, Claude Agent SDK, Semantic Kernel, or similar — and can walk us through the architecture and the failure modes you hit.
• You can speak fluently about the current state of agent orchestration: model differences across GPT, Claude, and Gemini; context engineering; tool-use patterns; evaluation strategy; guardrails; and cost/latency trade-offs — including how the frontier has moved in the last six months.
• You have shared your thinking publicly or internally — blog posts, OSS contributions, conference talks, internal write-ups, or a portfolio that shows how you reason about this space.
• Genuine passion and engagement with the work — obvious from the first conversation. You read the research, you try the tools the week they drop, you argue with your peers about where this is going.

• Experience with Microsoft Fabric, Power Platform, and Copilot Studio at enterprise scale.
• Working familiarity with additional model families beyond the big three — Mistral, Llama, Qwen, DeepSeek, or similar — and judgment on when open-weights make sense.
• Experience with eval frameworks (Promptfoo, Braintrust, DeepEval, LangSmith, OpenAI Evals, Anthropic’s eval tooling) or well-designed in-house approaches.
• Responsible AI and governance experience in regulated industries — healthcare, financial services, or public sector.
• Depth in one or more of: RAG at scale, fine-tuning / LoRA, vector search tuning, multi-modal systems.
• Consulting or professional-services experience, or prior work in a partner/ISV ecosystem.
• Track record of mentoring engineers or growing a practice.

Cyclotron is an Equal Opportunity Employer. Cyclotron values diversity, equity and inclusion, and aims to practice DE&I in all that we do.

Powered by JazzHR