Most engineering orgs bought AI coding tools and got nothing measurable back. The org that changes how it hires, reviews and validates gets 25% faster cycle time to PR. That's the job at CloudBolt: VP Engineering, 75+ engineers across the UK, US and India, running Claude Code adoption as a structured programme rather than a licence hand-out. Before that: Senior Director at Smarsh, building platform engineering, SRE and SaaS operations from scratch for a $150M ARR product line within a PE-backed compliance platform. Org 50 to 150+. $40M+ P&L. Extended exec team. Smarsh: 5 to 30+ enterprise deployments in 24 months, 99.9% uptime, $6M cut from annual cloud spend. CloudBolt: legacy monolith to event-driven AWS serverless, plus RAG over customer cloud inventory and cost data, with multi-tenant isolation, prompt-injection defence, output validation and agent traces on OpenTelemetry. Applied AI in production, not demos. The useful part is knowing where AI compounds capability and where it just compounds technical debt. I write about that at https://nimeshjm.com/blog Short version: Nimesh Manmohanlal builds AI-native engineering organisations for enterprise SaaS companies: the org design, not just the tooling. If you're a SaaS company hiring a VP or Head of Engineering to do this properly, message me here with the role and the constraint you can't get past.