• Enablement & coaching — work directly with engineers through 1:1s, pair programming, workshops, office hours, and cohorts to raise their effectiveness with AI-assisted development: prompting, workflow design, and model/tool selection. Identify under-adoption and lift teams up to the standard. • Best-practice propagation — capture what works (and what fails) in one team and systematically transfer it: living best-practice guides, prompt libraries, troubleshooting playbooks, and an internal AI knowledge base and skills registry. • Scale through champions — stand up and support a network of embedded AI champions so enablement scales past a single person. • Unblocking & feedback loop — be the go-to when engineers get stuck with AI tooling. Proactively tour teams to surface blockers early, aggregate them into patterns, and escalate systemic issues to leadership. • Internal integrations & tooling — build internal AI integrations and reusable tooling hands-on (Node.js/TypeScript, MCP) — from prototype to production independently, including auth, access control, security review, and CI/CD. • Evaluation & scouting — run structured evaluations of AI tools, models, and internal integrations; monitor the fast-moving landscape (Claude Code, Cursor, Copilot/Codex, new model releases) and deliver curated, actionable recommendations. • Measurement & impact — define and track adoption and productivity metrics by team, tool, and individual. Build dashboards on real usage and output — adoption/active-usage rate, assistant acceptance rate, PR throughput and cycle time — with a quality guardrail (change failure rate) so speed never comes at the cost of quality. Reference frameworks: DX Core 4, DORA. • Cross-functional partnering — partner with security/compliance on safe-by-design guardrails, and with GTM to translate field needs into internal enablement and feed adoption patterns back.

• Enablement & coaching — work directly with engineers through 1:1s, pair programming, workshops, office hours, and cohorts to raise their effectiveness with AI-assisted development: prompting, workflow design, and model/tool selection. Identify under-adoption and lift teams up to the standard. • Best-practice propagation — capture what works (and what fails) in one team and systematically transfer it: living best-practice guides, prompt libraries, troubleshooting playbooks, and an internal AI knowledge base and skills registry. • Scale through champions — stand up and support a network of embedded AI champions so enablement scales past a single person. • Unblocking & feedback loop — be the go-to when engineers get stuck with AI tooling. Proactively tour teams to surface blockers early, aggregate them into patterns, and escalate systemic issues to leadership. • Internal integrations & tooling — build internal AI integrations and reusable tooling hands-on (Node.js/TypeScript, MCP) — from prototype to production independently, including auth, access control, security review, and CI/CD. • Evaluation & scouting — run structured evaluations of AI tools, models, and internal integrations; monitor the fast-moving landscape (Claude Code, Cursor, Copilot/Codex, new model releases) and deliver curated, actionable recommendations. • Measurement & impact — define and track adoption and productivity metrics by team, tool, and individual. Build dashboards on real usage and output — adoption/active-usage rate, assistant acceptance rate, PR throughput and cycle time — with a quality guardrail (change failure rate) so speed never comes at the cost of quality. Reference frameworks: DX Core 4, DORA. • Cross-functional partnering — partner with security/compliance on safe-by-design guardrails, and with GTM to translate field needs into internal enablement and feed adoption patterns back.

• Enablement & coaching — work directly with engineers through 1:1s, pair programming, workshops, office hours, and cohorts to raise their effectiveness with AI-assisted development: prompting, workflow design, and model/tool selection. Identify under-adoption and lift teams up to the standard. • Best-practice propagation — capture what works (and what fails) in one team and systematically transfer it: living best-practice guides, prompt libraries, troubleshooting playbooks, and an internal AI knowledge base and skills registry. • Scale through champions — stand up and support a network of embedded AI champions so enablement scales past a single person. • Unblocking & feedback loop — be the go-to when engineers get stuck with AI tooling. Proactively tour teams to surface blockers early, aggregate them into patterns, and escalate systemic issues to leadership. • Internal integrations & tooling — build internal AI integrations and reusable tooling hands-on (Node.js/TypeScript, MCP) — from prototype to production independently, including auth, access control, security review, and CI/CD. • Evaluation & scouting — run structured evaluations of AI tools, models, and internal integrations; monitor the fast-moving landscape (Claude Code, Cursor, Copilot/Codex, new model releases) and deliver curated, actionable recommendations. • Measurement & impact — define and track adoption and productivity metrics by team, tool, and individual. Build dashboards on real usage and output — adoption/active-usage rate, assistant acceptance rate, PR throughput and cycle time — with a quality guardrail (change failure rate) so speed never comes at the cost of quality. Reference frameworks: DX Core 4, DORA. • Cross-functional partnering — partner with security/compliance on safe-by-design guardrails, and with GTM to translate field needs into internal enablement and feed adoption patterns back.

• Enablement & coaching — work directly with engineers through 1:1s, pair programming, workshops, office hours, and cohorts to raise their effectiveness with AI-assisted development: prompting, workflow design, and model/tool selection. Identify under-adoption and lift teams up to the standard. • Best-practice propagation — capture what works (and what fails) in one team and systematically transfer it: living best-practice guides, prompt libraries, troubleshooting playbooks, and an internal AI knowledge base and skills registry. • Scale through champions — stand up and support a network of embedded AI champions so enablement scales past a single person. • Unblocking & feedback loop — be the go-to when engineers get stuck with AI tooling. Proactively tour teams to surface blockers early, aggregate them into patterns, and escalate systemic issues to leadership. • Internal integrations & tooling — build internal AI integrations and reusable tooling hands-on (Node.js/TypeScript, MCP) — from prototype to production independently, including auth, access control, security review, and CI/CD. • Evaluation & scouting — run structured evaluations of AI tools, models, and internal integrations; monitor the fast-moving landscape (Claude Code, Cursor, Copilot/Codex, new model releases) and deliver curated, actionable recommendations. • Measurement & impact — define and track adoption and productivity metrics by team, tool, and individual. Build dashboards on real usage and output — adoption/active-usage rate, assistant acceptance rate, PR throughput and cycle time — with a quality guardrail (change failure rate) so speed never comes at the cost of quality. Reference frameworks: DX Core 4, DORA. • Cross-functional partnering — partner with security/compliance on safe-by-design guardrails, and with GTM to translate field needs into internal enablement and feed adoption patterns back.