Muhammad Saad Tanveer

Led a 4-person cross-functional team delivering React Native mobile apps and React/Next.js web applications, coordinating daily with Product, Backend, and Design. Owned architecture decisions and enforced code review standards for a multi-developer codebase, improving onboarding and maintainability. Resolved performance issues on data-heavy screens to improve end-user responsiveness. Designed an AI-assisted design-to-code workflow that reduced UI implementation time from about a week to a single day. Replaced an ad hoc delivery process with daily code pushes and ticket-based tracking to improve visibility and predictability, delivering across three concurrent applications.

Led a 4-person cross-functional team delivering React Native mobile apps and React/Next.js web applications. Owned architecture decisions and led code reviews, keeping the codebase clean and easy to onboard. Integrated applications with backend services and resolved performance issues to improve responsiveness on data-heavy screens. Designed an AI-assisted design-to-code workflow to reduce UI implementation time for new screens. Replaced ad-hoc processes with daily code pushes and ticket-based tracking for full delivery visibility. Delivered across three concurrent applications while balancing architecture, releases, and day-to-day development.

Owned platform-level delivery across a mobile portfolio serving tens of millions of program beneficiaries, building internal systems and re-architecting a business-critical field application. Built a centralized internal platform for organizational queries, tasks, and sensitive health data with role-based access controls, audit trails, and controlled file sharing. Modeled reporting hierarchy so tasks and approvals route correctly along reporting lines, eliminating informal tracking and improving auditability. Re-architected the field patient application to an offline-first approach with local-first sync and conflict resolution. Deployed on-device vector and audio recognition ML pipelines on iOS and Android, reducing inference latency by 35%. Shipped production document generation supporting image-only, text-only, and combined outputs, and mentored developers on secure data handling and offline-first patterns.