As one of the largest energy suppliers in the world, bp offer a fuel card program to businesses for their fleets to fuel & charge vehicles at discounted rates. Their portal to manage transactions was built in the 90's and they needed support to provide a newer platform to customers to avoid further market share decline in this sector.

What shipped

• Fleet Manager Portal
• Fuel Card Onboarding Forms

Impact

• Time to market: full portal delivered in 3 months (previous attempt took 3 years).
• Performance: ~80% uplift on key journeys; faster pages improved migration and NPS.
• Customer recovery: modern UX stemmed further market-share decline in the fleet segment.
• Sustainability: contributed to >60% improvement across the fleet digital estate via page-weight and runtime optimisations.

My role & scope

• Led 6+ engineering teams across FE/BE/QA to sequence delivery, de-risk launches, and hit dates.
• Owned front-end architecture & standards; introduced performance budgets and a11y checks in CI.
• Drove strangler-pattern migration off legacy, coordinating with product, security, and markets.
• Set up release runbooks, feature flags, and rollback paths; coached engineers and tech leads.

Key decisions & trade-offs

• Rebuild vs patch legacy: opted for a greenfield front-end with progressive cutover to minimise risk.
• Service layer: pragmatic Node + .NET mix to reuse proven capabilities while enabling faster FE delivery.
• Platform choice: AWS (CDK) for hosting and delivery, releasing with Azure DevOps pipelines.
• Quality gates: enforced perf/a11y budgets in CI over chasing perfect lab scores. Balanced speed with standards.

See it in action

Tech Stack

React, TypeScript, Node.js, .NET C#, AWS CDK, AWS (Lambda, CloudFront, S3), Azure, Azure DevOps