Arpit Agarwal

Built the live-response layer for an embeddable AI assistant, implementing token streaming over WebSocket with a state machine and safe rendering of model-driven UI. Implemented source-verifiability UX with citations and a sources modal.

Led and mentored a team of 5+ engineers on a B2C insurance platform, owning business-critical modules and driving AI-assisted development. Drove performance and reliability improvements, cutting load time by ~35% and building observability dashboards.

Owned frontend development as the first engineer, shaping product vision and architecture for a log-based observability platform. Built core UI including log parsing, event-sequence builders, and interactive charts from streaming data.

Built and maintained agent portals, improving accessibility and operational efficiency. Set up a widgetization workflow and a Chrome extension for WhatsApp Web analytics.

Led frontend development for the Vega team, architecting reusable components and shipping key features. Drove code quality and technical-debt reduction, cutting technical debt by 32% and improving application speed by 20%.

Designed and built customer-health dashboards for the People Management team, surfacing NPS and engagement data to reduce churn. Architected the Influencer and Hierarchical Map features, which became high revenue-generating modules.

Delivered web application features for Sprint Telecommunications, building responsive UI screens and reusable components. Resolved production defects and upheld quality across the software delivery lifecycle.

Pursued a Bachelor of Technology in Information Technology from 2011 to 2015.