• Hands-on daily fluency with agentic coding stacks: Claude Code, Cursor, Codex — including MCP servers, skills, sub-agents, and spec-driven development on real repositories
• AI architecture: LLM gateways/proxies (LiteLLM / OpenRouter class), cost and error attribution, local-LLM trade-offs, in-region deployment patterns (e.g. Bedrock)
• Engineering-leadership credibility at tech-lead / AI-architect / head-of-engineering level — able to review real code, pipelines and specs with senior engineers, not present slides
• Facilitation of technical sessions: live demos, validation calls, hands-on workshops with real repos
• Packaging: turning a working practice into an artifact another team adopts without the author in the room
• Knowledge
• Engineering measurement in the AI era: DX Core 4 / DORA, AI-impact metrics, quality guardrails for AI-generated code
• Adoption psychology for senior engineers — resistance among seniors is a named blocker in several of the client's companies
• Game development / iGaming exposure (nice to have): game math, certification constraints, art/animation pipelines
• Led AI adoption or platform/developer-enablement work in an engineering organization (20+ engineers), or equivalent tech-lead/head-of-engineering experience
• Shipped agentic workflows to production; can show their own skills, MCP servers or spec repositories
• Fluent English required; Russian and/or Ukrainian a strong plus — the client's teams communicate in both