Jitendra Kumawat

• Led platform-wide frontend optimisation, cutting shared first-load JavaScript from 1.03 MB to 262 kB (75%), production build time from 228s to 104s (54%), first-load CSS by 81% and deployment weight by 88 MB through dependency pruning, modular architecture, import audits and tree-shaking.

• Built and scaled Forge, a Slack-native agentic AI engineering platform adopted by 27 colleagues across six functions and 15 Slack channels in its first three months. It enabled 110+ human-reviewed merged pull requests and resolved 900+ codebase and analytics requests.

• Built Forge's control plane: service health dashboards, incident detection, release and deployment tracking, audit trails, and review-gated agent and skill updates.

• Redesigned the partner application-filing workflow to 99.8% adoption within a week, cutting draft abandonment from 8.3% to 2.6% with median filing time under a minute.

• Built the partner portal's primary action dashboard, used by ~88% of 5,300 monthly partner use

• Rebuilt and migrated Course Finder across two applications without a release freeze. It was adopted by 3,324 monthly partners (~63% of portal users) and generated 91,000+ filter actions and 85,000+ course shortlists in two months.

• Owned the MUI and Tailwind design system, introducing token-based theming and reusable components.

• Built the Rules Management Engine UI from scratch, with recursive condition groups for complex email and workflow automation rules.

• Built an AI-powered visa mock interview platform with React, LiveKit, Python and OpenAI. It ran in production for about 14 months, peaked near 735 sessions a week, and held a 4.1/5 rating.

• Awarded Excellent Performer within four months.