K. Jagadish

Agentic AI Development & LLM Integration − Agentic Coding Operating Model: Runs a production agentic development workflow orchestrating Claude Code, OpenCode, and Cursor against multiple LLM providers (Claude, GPT, GLM), with Model Context Protocol (MCP) servers — Context7, Figma, Playwright, and Sentry — wired into the loop to give agents live access to library documentation, design specs, browser automation, and production error telemetry. − Complex Feature Delivery via Agentic Workflow: Designed and shipped a drug experimentation project management module — coupling experiment tracking to live inventory consumption and the material lineage graph, a combination with no direct off-the-shelf equivalent — using an agent-driven workflow to compress specification, implementation, and validation cycles on a problem with no reference implementation to draw from. − LLM-Powered Operational Automation: Integrated Azure OpenAI Service into core platform operations, delivering automated summarization and intelligent triage across workflows, inventory movements, and operational events, reducing manual review load on business teams. − Schema-First Contract-Driven Pipeline: Architected a contract-first workflow converting GraphQL schemas and Swagger/OpenAPI specifications into auto-generated mock services, Zod validation schemas, TypeScript types, and React Query hooks; enabled parallel UI development ahead of backend implementation, cutting live integration timelines to 1 day. − End-to-End Ownership Under Ambiguity: Owns technical direction from problem framing through production delivery as founding technical leader, rapidly pivoting architecture in response to shifting customer requirements and market feedback without external structural support. − Enterprise SaaS Platform Delivery: Built Traceflow, a standalone Next.js enterprise platform with a design system layered on Material UI, engineering custom high-performance data tables that render and manipulate datasets exceeding 30,000 items, powering revenue-loss discovery and recovery workflows. Data Visualization & Performance Engineering − Custom Spatial Virtualization Engine: Engineered a graph visualization system on ReactFlow to bypass native rendering limits at ~1,000 nodes; built a 9-grid spatial viewport virtualization algorithm with predictive node pre- loading and off-screen unloading, sustaining 60 FPS panning and zooming across supply chain topologies with 800+ complex nodes. − Interactive BOM Calculation Engine: Built a Bill of Materials module inside the graph interface supporting real-time multi-level yield calculations and dynamic unit conversions across upstream and downstream nodes during product experimentation. Cloud Infrastructure, Security & Engineering Standards − Zero-Downtime Deployment Pipeline: Designed full-stack infrastructure using a Node.js/Next.js BFF layer, custom Docker containers, PM2 hot-cold instances, and automated Azure Front Door traffic swapping for zero- downtime releases. − Shift-Left Quality Governance: Established pre-commit hooks and CI/CD pipelines running Jest, React Testing Library, Cypress, and Playwright suites, catching regressions locally before they reach production. − Client-Side ABAC Framework: Engineered Attribute-Based Access Control by resolving custom JWT claims from Auth0 to dynamically gate UI routes, action controls, and granular permissions across tenant roles. − Multi-Tenant Marketplace Framework: Architected the foundational UI framework for a secondary vendor marketplace using Shadcn UI, Tailwind CSS, React Router, and Casdoor, handling multi-vendor onboarding, bidding, and contract negotiation.

Global Engineering Leadership & Standards − Distributed Chapter Governance: Led a multi-region, remote-first Frontend Chapter aligning developers across global time zones; defined technical debt roadmaps, drove TypeScript adoption, and automated PR and pre-commit quality checks. − Cross-Functional Domain Integration: Embedded into high-priority engineering teams as a staff-level contributor to unblock delivery bottlenecks, govern code reviews, and establish scalable UI architecture patterns. − Navigation Architecture Revamp: Co-led engineering execution for a full platform navigation and UI overhaul, partnering with Product and UX on technical feasibility and effort estimation. Architecture & Legacy Modernization − Node.js BFF Layer: Built and maintained Node.js backend-for-frontend services mediating between the React client and backend domain services, consolidating fragmented API surfaces into UI-shaped contracts. − Contract-First API Workflows: Integrated automated entity generation from Swagger definitions into the frontend pipeline, establishing a contract-first paradigm that streamlined API integration and improved cross-stack alignment with backend teams. − Legacy Decoupling & Codebase Unification: Modernized a fragmented legacy frontend by systematically deprecating Backbone.js files and adapter layers, migrating features to a unified React component library and eliminating adapter-induced code bloat. − Monolith Infrastructure Decoupling: Redesigned application configuration architecture into a lean modular setup, isolating legacy frontend assets from a monolithic backend.

Distributed Architecture & Platform Leadership − Unified Micro-Frontend Framework: Re-engineered 8 discrete, subdomain-isolated SaaS applications into a unified micro-frontend architecture using Webpack Module Federation, routing applications under clean sub-paths beneath a central shell container. − Shared Platform Services: Centralized global user sessions, permissions, dynamic configuration, and plan entitlements in the shell application, eliminating duplicate logic and redundant API calls across app modules. − Inner-Sourcing & Platform Leadership: Led the Frontend Platform Engineering team on a rotating contribution model, empowering cross-functional engineers to contribute reusable components and share ownership of platform infrastructure. Performance, Design Systems & Cloud Delivery − Latency Optimization & Network Routing: Re-architected API routing through internal API Gateways and optimized pod configuration services instead of public endpoints, significantly cutting network latency across frontend microservices. − Enterprise Component Library: Architected the default enterprise component library and boilerplate framework using React, TypeScript, and Ant Design, documented in Storybook for all current and future platform development. − Static Cloud Hosting Pipeline: Streamlined deployment infrastructure using AWS S3 and CloudFront CDN via API Gateway to deliver high-availability, low-latency assets to enterprise clients.

− Predictive AI SaaS Platform: Contributed to the core frontend of bioplan.ai, an AI-driven aquaculture production planning platform used by fish farming operations across Norway. − Model Output Interfaces: Designed data-dense analytics dashboards translating complex optimization model outputs into actionable telemetry — biomass growth predictions, feed consumption curves, and environmental metrics.

Machine Learning Platform Engineering − ML Lifecycle Management Platform: Engineered an enterprise ML model lifecycle platform using Ember.js and Python (Flask), enabling data scientists across LinkedIn to track model performance, automate artifact versioning, and streamline model deployments. − Developer API & Dataset Discoverability: Built an interactive API sandbox and dataset discovery suite that reduced redundant development effort across global internal engineering teams. − Custom Component Architecture: Re-engineered design system components and UI controls from scratch in Ember.js to support complex, high-density data management workflows. Performance & CI/CD Governance − Automated CI/CD Quality Metrics: Extended LinkedIn's internal CI/CD toolchain to automatically compute code health metrics and run automated code review validations during the pull request process. − Render Performance & TTI Optimization: Improved Time-to-Interactive and render speed across internal web apps by standardizing rendering patterns and eliminating heavy legacy dependencies.

Distributed Frontend Architecture − Event-Driven Shell Architecture: Formed part of the founding Frontend Chapter for Sam's Club, architecting an early shell-based container hosting multiple independent UI applications communicating via an in-browser event bus, predating modern micro-frontend tooling, backed by Node.js BFF services aggregating data across supply chain backend systems. − Angular-to-React Modernization: Evaluated, benchmarked, and authored the technical standards for migrating legacy AngularJS 1.6 applications to React with comprehensive automated test coverage. Data Platform & Forecasting Tools − Smart Forecasting Platform: Architected the UI for an internal demand forecasting tool used by supply chain managers to analyze historical sales trends, manage procurement orders, and optimize inventory replenishment across stores. − Internal Data Lake & ETL Tooling: Built the web interface for an in-house Data Lake, engineering configurable visualization components, dataset exploration tools, and interactive ETL pipeline builders. − Identity & Location Analytics: Developed a "Golden Profile" visualization tracking unified customer identity across omnichannel touchpoints and contributed to a spatial planning tool predicting store cannibalization and population reach. − Technical Mentorship: Mentored returning-to-work women engineers in building automated end-to-end Cypress test suites, strengthening release stability.

Glider.ai — Data-Science-Driven Talent Platform − Real-Time Reactive Architecture: Led an engineering team building Glider.ai, a real-time, data-science-driven candidate evaluation SaaS platform on Meteor.js, Node.js, and MongoDB, delivering live candidate-recruiter matching, messaging, and automated assessments. − NLP Ingestion Pipelines: Integrated Python-based document parsing and NLP engines to ingest multi-source resumes and job profiles, powering dynamic job-matching recommendations. Edulastic — EdTech Learning Platform − UI Modernization & Accessibility: Overhauled the core Backbone.js, Highcharts, and jQuery UI interface, refactoring legacy components for mobile responsiveness and web accessibility compliance. − Searchable Content Interface: Designed an interactive search interface enabling educators to query and index learning materials efficiently.

− Flagship Platform Overhaul: Worked in a 5-member core engineering team to rebuild zap2it.com, establishing the frontend UI framework and local development environments using YUI3 and FuelPHP. − Solr Search Integration: Integrated Apache Solr's API cached on Varnish, to power site-wide queries across cast, crew, TV schedules, movie metadata, and editorial content. Implemented a cache warming service to allow editors to pre-cache wildly popular content like articles related to trending shows like, Game Of Thrones, Arrow etc. − Unified Analytics & Monetization Framework: Implemented a reusable analytics framework integrating Adobe Test & Target, Google Analytics, Chartbeat, and Nielsen to drive content personalization and ad monetization. − Third-Party Content Ingestion: Built dynamic modules to ingest, parse, and render live syndicated news content from multiple media providers.

− Seller Onboarding Platform: Engineered high-concurrency Node.js services and frontend interfaces (Backbone.js, AngularJS) for the seller onboarding platform underpinning Flipkart's pivot to third-party sellers. − Complex Distributed Integrations: Built the primary Order Management UI, orchestrating asynchronous workflows across backend microservices including SwiftMQ message brokers, catalog management, dynamic pricing, and delivery scheduling, through a Node.js BFF layer that aggregated and normalized responses for the order management client. − Peak Season Performance Optimization: Diagnosed and resolved order processing bottlenecks ahead of high-traffic festive sales, establishing resilient asynchronous processing patterns for extreme traffic spikes. − Developer Support Tooling: Built a real-time order tracking dashboard enabling support teams to isolate and resolve stuck orders across the fulfillment pipeline.

Test & Target Developer − Sub-100ms Script Execution: Built high-performance, DOM-injected JavaScript campaigns via Adobe Test & Target across global Dell web domains, executing under 100ms to eliminate layout shift and render delay. − Conversion Optimization: Engineered personalized recommendation logic and checkout flow optimizations, converting winning experiments into permanent production features.