Johnson Chinonso

• Building a DevOps co-pilot. React and TypeScript on the frontend, Node.js and GraphQL on the backend, Postgres underneath. Multi-tenant from day one. • Built out the CI/CD log analysis backend: ingestion endpoints, a confidence scoring engine, and a risk classification layer that feeds failure reports back to the UI as runs come in. • Designed the RBAC system top to bottom: data model, API, and the UI for managing it. Customers can scope what the co-pilot sees and acts on across roles and environments. • Iterated heavily on the prompt layer to keep LLM output anchored to actual log evidence instead of confident- sounding inference. Hallucination rate dropped a lot in early testing. • Stood up CI/CD, API scaffolding, and the component library from zero. Small team, fast pace, wearing a lot of hats.

• Set up the frontend foundation and design system in React, Next.js, and TypeScript. The code standards and component library held up as the team grew from 2 to 7 engineers. • Led frontend on the SpinBet casino platform, including Tournament Platform, Campaign System, and a real-time Bet Ticker. Daily actives went up around 22% and user complaints dropped roughly a third. • Spent a sprint chasing down LCP and CLS regressions on the main flows. Got LCP from around 4s to under 2s, and CLS down to about 0.05. Mostly code splitting, image work, and lazy loading. • Mobile session duration climbed about 30% after a responsive overhaul and moving more pages onto Next.js SSR. Desktop conversion picked up too. • Shipped 40+ accessible UI components with TypeScript types and proper Storybook docs. New engineers got productive faster after that. • Ran Lighthouse audits and brought customer-facing pages up to WCAG 2.1. Accessibility scores landed in the mid-90s. • Worked with the backend team on the migration from the legacy API to the new architecture. Zero downtime for around 50K active users.

• Built core modules for ScienceLogic's SL1 SaaS platform (React, TypeScript, GraphQL). The platform monitors 100K+ devices across enterprise customers. • Took the SLDATA-Grid component from around 3s load to about 1.4s. Selective GraphQL queries, virtualized rendering, and some frontend caching. • Designed and shipped the Device Groups feature using a component-driven approach, which cut build time on similar features roughly in half. • Wrote the frontend code quality and performance guidelines the team used in sprint reviews. New engineer onboarding dropped from about a month to two weeks. • Worked closely with Product, UX, and Backend in Scrum across six major releases.

• Led a team of seven engineers on the Africlick mobile app (React Native, TypeScript). Code reviews, pairing sessions, and a fair bit of mentoring. • Got the React Native codebase to a consistent 60fps on both Android and iOS, which moved user retention by about a third. • Set up the engineering standards: TypeScript strict mode, mandatory tests on PRs, PR templates, CI/CD. Post-deploy bug count came down a lot. • Tracked down memory leaks that were causing roughly 15% of app crashes. Mostly listener and image-cache issues.

• Built the Category Mapper, which automated merchant product classification for Google Shopping, Amazon, and Shopify integrations. Cut manual entry by more than half. • Rebuilt the web client in React and Gatsby. Page loads went from around 5s to roughly 3s after the static generation and image work. • Worked on backend APIs handling north of a million requests an hour. Caching layers and a few query rewrites got uptime to 99.9%.

• Built the Sechub automotive platform frontend from scratch in React and React Native. • Worked with the design team on a UI/UX redesign. User adoption and retention both improved after release. • Set up the cross-platform testing approach with Jest and end-to-end tests.

• Built REST APIs in Node.js and Express for the Voice of Andela platform. Handled 500K+ concurrent users with sub-200ms responses and full unit test coverage.

• Worked on the Open Heaven mobile products in React and React Native. App store ratings went up after a round of UI changes driven by user feedback. • Added proper error handling and performance monitoring, which knocked out a chunk of API-related bugs.