Case study
National Grid: a cloud and AI platform for a siloed enterprise
Owned the cloud and AI/ML roadmap at a 20,000-person utility: launched a RAG-based generative AI platform to 3,000+ monthly users and grew an internal cloud services platform from $2M to $20M+ in annual chargeback.
Confidentiality. Details are under NDA; this describes my role and approach, not proprietary specifics.
Context
National Grid is a 20,000-person Fortune 500 utility. From 2022 to 2025 I was a product manager there, promoted from Technical to Senior to Principal Product Manager. I owned the strategic cloud and AI/ML roadmap.
The problem
Engineering was organized in silos. Each business unit ran its own infrastructure, and adopting a tool meant a separate procurement and security review for every team. Deploying a cloud service took weeks. Engineers and support teams also needed faster answers from internal documentation. Product management was new to much of the organization; the data science team saw it as overhead.
My role
Principal Product Manager. I set the cloud and AI/ML roadmap, prototyped AI systems alongside engineers, explained technical strategy and business value to the C-suite, and turned the strategy into engineering plans. I grew the team from 1 to 4 pods and hired and developed 4 product managers.
What I built / did
- Designed and launched an internal generative AI platform that connected an LLM (GPT) to proprietary documentation through retrieval-augmented generation (RAG), so engineers and support teams resolved issues faster.
- Built an Infrastructure as Code (IaC) service and standardized 100+ enterprise tools (AKS, Snowflake, Power BI) across all business units, with no separate procurement or security review per team.
- Founded Cloud Leadership Sync, the company's first executive and Chief Architect-level cross-business-unit forum, and introduced portfolio-level roadmapping, PRDs and agile ceremonies across all business units.
- Ran individual stakeholder campaigns to get every business unit President and Chief Architect to give up infrastructure autonomy. Presidents heard about removing an engineering tax; Chief Architects heard about production-grade tooling without a security review burden.
Results
- Grew an internal cloud services platform from $2M to $20M+ in annual chargeback.
- The generative AI platform reached 3,000+ monthly active users.
- Cloud service deployment went from weeks to minutes.
- The engineering organization grew from 5 to 35 engineers.
- 2 of the 4 product managers I hired were promoted to senior within 12 months.
How I worked
Most of the job was persuasion without authority. I pitched each audience on its own problem rather than giving one deck to everyone. I won over the data science team through embedded discovery and their own data, until their view moved from "product management is overhead" to "product management is why our models get used." The documents were the standard product set, applied at portfolio scale: PRDs, portfolio roadmaps, a recurring executive forum, and engineering plans translated from strategy.