Senior Frontend Engineer · React, TypeScript, Next.js
I ship interfaces that survive real data: marketing intelligence dashboards, microfrontend platforms for 50k+ daily users, and agentic AI products. Based in Bangalore · open to remote-friendly roles.
Senior Software Engineer at Lifesight
Bangalore, India
7+ years · SaaS & GovTech
// about
I'm a Senior Frontend Engineer based in Bangalore, India, with 7+ years of experience building scalable SaaS platforms across analytics, automotive retail, and geospatial domains.
At Lifesight, I lead frontend development for a unified marketing intelligence platform, building the dashboards and AI-powered interfaces that help enterprise brands make smarter decisions. Before that, I spent 4+ years at Tekion scaling frontend systems used by 50,000+ daily users.
I believe great frontend engineering lives at the intersection of technical rigour and product empathy. I care about the milliseconds, the accessibility, the test coverage, not just the code that ships.
// common questions
Yes, currently open to full-time senior or staff frontend engineering roles and consulting engagements. Based in Bangalore, open to remote globally. Reach me at animeshpandey1909@gmail.com or LinkedIn.
// experience
Leading frontend development for Lifesight's Unified Measurement OS, a marketing intelligence SaaS combining MMM, incrementality testing, and causal attribution. Building SSR analytics dashboards, contributing to Mia (their agentic AI product), and establishing platform-wide testing and performance standards.
Owned 10+ mission-critical frontend modules across enterprise automotive retail. Built Storybook-documented component libraries and microfrontend architectures with Webpack Module Federation. Reduced load times by 30% for 50k+ daily users. Introduced AI-driven development tooling cutting team toil by 30%.
Developed enterprise automotive retail applications using ReactJS, Redux, and React Native. Built 20+ reusable UI components and microfrontend modules documented in Storybook, accelerating feature delivery across engineering teams.
Led frontend development of Kerala-WRIS dashboards providing real-time water resource visibility to millions of residents. Built geospatial visualisation systems for government clients and reduced API overhead by 30% through frontend optimisation.
Developed a real-time SCADA monitoring system for Andhra Pradesh Water Resources using AngularJS and D3.js, enabling operational dashboards for field teams.
// skills
Production tools from recent platforms - not a buzzword dump.
// projects
Selected work across SaaS analytics, automotive retail, and GovTech.
Architecture · Tekion
Architected and shipped a production microfrontend system using Webpack Module Federation across 10+ mission-critical enterprise automotive retail modules, each owned by an independent squad and deployed without cross-team coordination. Achieved a 30% reduction in load times through aggressive code splitting, route-level lazy loading, and bundle-level Lighthouse budgets enforced in CI. Introduced AI-driven development tooling that cut engineering toil by 30%. The platform serves 50,000+ daily active users across dealership and OEM workflows.
GovTech · Vassar Labs
Led frontend development for the Government of Kerala's statewide water resource information system. Built real-time geospatial dashboards over Leaflet with district-level overlays, time-series hydrology charts in Highcharts, and D3.js data-layer visualisations for field operations teams. Reduced API overhead by 30% through frontend caching and request batching strategies. Concurrently built a real-time SCADA monitoring system for Andhra Pradesh Water Resources used by field teams for operational visibility.
AI Product · Lifesight
Built the frontend for Lifesight's agentic AI assistant, which interprets complex marketing datasets and surfaces attribution insights through a streaming conversational interface. Implemented real-time LLM response streaming, tool-call visualisation panels that expose reasoning steps, and deep integration with the Unified Measurement OS data layer. Designed to handle ambiguous multi-step queries from non-technical marketing stakeholders with progressive disclosure of confidence and caveats.
AI Feature · Lifesight
Embedded AI layer within the Lifesight attribution dashboard that continuously monitors media spend signals and proactively surfaces anomalies without requiring the user to go looking. Statistical changepoint detection (z-score + PELT algorithm) runs on the data pipeline; results are handed to an LLM that generates plain-English explanations of what shifted, why it likely happened, and what action to consider. Surfaces budget pacing issues, channel-level attribution drift, and incrementality test failures as inline insight cards - not alerts that get ignored.
Analytics Platform · Lifesight
Delivered enterprise-grade analytics dashboards powering Lifesight's Unified Measurement OS: MMM (Media Mix Modelling) curves, incrementality test results, and causal attribution breakdowns for enterprise brands managing multi-million-dollar ad budgets. Built with Next.js SSR for fast first-paint on dense data-heavy views, enforced Core Web Vitals budgets in CI, and established platform-wide Playwright testing standards that became the team baseline.
Design System · Tekion
Designed and maintained a Storybook-documented shared component library used across 10+ product modules at Tekion. Built 20+ reusable, accessible UI components (forms, data tables, modal flows, navigation patterns) with full TypeScript typing, WCAG 2.1 compliance, and interactive documentation. Reduced per-team duplicate code by 25% and measurably shortened time-to-feature across multiple engineering squads.
AI Tooling · Tekion
Internal CLI and VS Code extension that automated the highest-toil parts of Tekion's frontend development cycle, cutting team overhead by 30%. The scaffolder generates a fully wired microfrontend module in under 10 seconds - Webpack Module Federation config, Storybook stories, Jest setup, and GitHub Actions job - from a single command. An AI-assisted PR review workflow uses GPT-4 to flag accessibility regressions, performance anti-patterns, and missing test coverage before human review. Also shipped automated migration scripts for the React 16 to 18 upgrade across all 10+ modules, completing the migration in days rather than weeks.
Mobile · Tekion
React Native companion to Tekion's web-based dealer management platform, designed for dealership staff who work on the shop floor and can't stay at a desk. Covers real-time inventory lookup with VIN scanning, live service queue visibility, and customer CRM access. Built offline-first with SQLite-backed local state that syncs over a conflict-free delta protocol when connectivity restores - critical in underground service bays with patchy signal. Deep-linked from the web platform and shares Redux slices with the web app via a shared library, keeping business logic in one place.
Developer Tooling · Tekion
GitHub Actions pipeline that runs Lighthouse CI against a Playwright-rendered preview deployment on every pull request, automatically blocking merges when Core Web Vitals or bundle size regresses beyond configured thresholds. Posts a generated PR comment with before/after metric tables, a bundle diff, and a 30-day trend sparkline so reviewers have full context without leaving GitHub. Historical performance data is persisted to a Postgres store and surfaced on an internal dashboard used by engineering leadership to track platform health across all 10+ microfrontend modules. Went from no performance visibility to catching 12 regressions in the first month.
// writing
I write about the JavaScript patterns, system design decisions, and frontend engineering trade-offs I encounter at work - the stuff that doesn't make it into the docs.
// case studies
31 live, scrollable frontend case studies — skeleton screens, perceived performance, streaming UIs. With interactive demos and reading levels for junior, mid, and staff engineers.
Explore the CasebookWhat makes a function pure, why procedures aren't functions, and how declarative thinking transforms the way you write JavaScript. A primer on functional programming fundamentals.
this?
JavaScript's this is a runtime binding determined by call-site context. This piece works through all four binding rules (new, explicit, implicit, default) in order of precedence.
How to build a streaming AI agent UI that feels native to your product. SSE lifecycle, tool-call panels, AbortController, UX states for thinking / streaming / error - and why skipping the chat bubble is the right call.
this in JavaScript
An in-depth look at JavaScript's most misunderstood keyword: implicit, explicit, and new binding rules demystified with real examples.
Essays in progress — topics from production frontend engineering.
Coming Q3 2026
How we wired Lighthouse CI, bundle budgets, and PR comments so 10+ microfrontends couldn't regress Core Web Vitals silently.
Coming Q4 2026
Module Federation worked until it didn't - versioning, shared dependencies, and the debugging stories nobody puts in the sales deck.
Researching
Patterns from MMM and attribution UIs: hierarchy, defaults, and when not to add another chart.
Coming 2026
The migration playbook we used at Tekion - codemods, feature flags, and the tests that caught what types missed.
Exploring
Why I built a briefing mode, what I'd do differently, and how static sites can still feel product-grade.
Backlog
When particles earn their keep - and when prefers-reduced-motion should win.
Backlog
Tokens, lazy features, and discipline - how this portfolio is structured and why.
// education
B.Tech in Computer Science and Engineering
// contact