Jon Reader

R&D leadership

Ran R&D across 8 teams (7 server-rendered apps and an internal CMS): trialled new tools in short spikes, such as AWS chaos testing, shared what worked in brown-bag sessions and rolled it out.

• AWS Fault Injection Service
• AWS Lambda
• AWS CloudWatch

AI engineering platform

Built shared GitHub Actions workflows that let AI agents write code themselves and review pull requests across Node.js, Python and Java repositories. Set models (Claude, GPT) centrally and added guardrails that limit which tools and commands agents can run, with short-lived OIDC credentials instead of stored keys.

• GitHub Actions
• Augment
• Claude
• OpenAI GPT
• Node.js
• Python
• OIDC
• AWS Secrets Manager

Usher, internal AI assistant

Contributed to an internal chatbot that answers questions and triages issues in real time by connecting to other internal systems through MCP, and integrated it with our Agentic Platform. Usher runs on Next.js, the Vercel AI SDK and Amazon Bedrock.

• Next.js
• TypeScript
• Vercel AI SDK
• Amazon Bedrock
• AWS DynamoDB
• AWS S3

Agent skills

Wrote a range of Cursor agent skills, for example one that triages incidents by pulling Datadog, GitHub and Slack evidence into a single timeline, and one that reviews components across repositories.

• Cursor
• Python
• Datadog

AI developer tools

Created an MCP server that gives AI agents access to internal APIs, a Datadog extension for VS Code and a documentation-migration plugin that waits for human approval before making changes.

• Model Context Protocol
• AWS Lambda
• Datadog
• VS Code

Multi-brand design system

Led the end-to-end migration of the React 19 toolkit shared by Peacock, NOW, Sky (100+ components) from styled-components to Tailwind CSS v4. Built an AI snapshot-diff tool so agents could catch and fix visual regressions in migrated components without a person in the loop.

• React 19
• TypeScript
• Tailwind CSS v4
• Styled Components
• Storybook
• Playwright
• AWS S3

CI and deployment

Rebuilt monorepo CI to test and build each package in parallel, rolled out shared GitHub Actions workflows to other repositories and introduced AWS Lambda Web Adapter to the organisation's apps through the Serverless Framework.

• GitHub Actions
• Concourse
• Nx
• AWS Lambda
• Serverless
• Playwright

Migrations and enablement

Drove organisation-wide migrations, including a central shared configuration, Node.js 24 and Vitest 4, and wrote guidance for AI agents (AGENTS.md) and new joiners across many repositories.

• Node.js
• TypeScript
• Vitest

Technical leadership

Wrote code day to day and took on the Technical Lead role on top: hired and mentored engineers, set the technical direction for the team's applications and was the main contact for Principal Engineers and stakeholders.

Feature delivery

Delivered features and day-to-day work across a React micro-frontend architecture, with GraphQL and REST APIs behind a backend-for-frontend (BFF). Also contributed shared components used across the wider NOW platform.

• React
• Next.js
• TypeScript
• GraphQL
• REST
• Node.js

Quality and accessibility

Applications carried 1,000+ unit, functional and end-to-end tests written against acceptance criteria, and nothing merged with an accessibility failure.

• Jest
• React Testing Library
• Cypress
• TypeScript

DevOps and on-call

Ran CI/CD, monitoring and on-call for the team's apps on Akamai, under a DevOps model.

• Concourse
• Akamai