Lead Frontend Engineer

Ford Business Solutions; Chennai; On-site; INR 1,200,000 - 1,800,000; Full time

Job summary

Ford Business Solutions seeks an experienced Frontend Architect to own the AI/Platform UI stack. The role focuses on defining patterns for streaming UI, real-time responses, and tool visualization, while enforcing a scalable design system and component library across product squads.

You will drive secure API integrations, performance budgets, accessibility conformance, and observability, collaborating with backend teams to deliver end-to-end, self-service frontend capabilities for the enterprise.

Qualifications

• Experience in leading frontend architecture for AI/Platform UI
• Ability to define and enforce design system and component architecture
• Proficiency with Next.js/React and modern state-management
• Experience integrating frontend with backend services (APIs, WebSocket, SSE)
• Strong focus on performance, accessibility (WCAG 2.1 AA) and observability
• Knowledge of security, auth flows (OAuth2) and DevEx tooling
• Experience shipping platform self-service UIs and SDKs

Responsibilities

• Own end-to-end frontend architecture for AI/Non-AI platform products.
• Define and enforce frontend tech stack, patterns, and state-management across UI surfaces.
• Architect integration layers between frontend and AI backends via secure APIs.
• Drive rendering strategies (SSR/CSR/ISR) and accessibility budgets.
• Build and maintain a reusable AI UI component library and design system.
• Create golden-path templates and starter kits for rapid AI UI deployment.
• Oversee platform service catalog frontend work and self-service provisioning UIs.
• Collaborate with infra to expose model routing and cost-aware choices through UI controls.

Skills

Frontend architecture

Design system

React/Next.js

API integration

Performance optimization

Accessibility WCAG

Observability

Tools

TypeScript

Storybook

REST/GraphQL

WebSocket/SSE

React.js

Node.js

Job description

Job Summary

• Own end-to-end frontend architecture for AI and Non-AI Platform products, defining patterns for streaming UI, real-time agent responses, tool-call visualization, citations/grounding display, and multi-turn conversational state.
• Define and enforce the frontend technology stack, coding standards, component architecture (design system), and state-management strategy across all AI Platform UI surfaces.
• Architect the integration layer between frontend applications and AI backend services (Vertex AI / Gemini, Google ADK agents, RAG pipelines, LangGraph/LangChain orchestrators) via secure, well-documented APIs (Fast/REST/GraphQL/WebSocket/SSE).
• Drive technical decisions on rendering strategy (SSR/CSR/ISR), edge/streaming delivery, performance budgets, and accessibility (WCAG 2.1 AA) for AI-heavy, latency-sensitive interfaces.
• Design and maintain the AI Platform's internal Developer Portal / Self-Service Console, enabling engineering teams to provision AI resources (RAG pipelines, agents, vector indices, environments) without manual tickets - extending self-service cloud infrastructure provisioning patterns into the UI layer.
• Build and version a reusable AI-UI Component Library / Design System (chat widgets, agent trace viewers, citation cards, feedback/human-in-the-loop widgets, prompt/response inspectors) consumable by multiple product squads across the enterprise as a shared platform capability.
• Define 'golden path' frontend templates and starter kits (Next.js/React + AI SDK boilerplate) so downstream product teams can spin up new AI-powered UIs in days, not weeks - a core platform-engineering productivity outcome.
• Own the frontend layer of platform Service Catalog entries (e.g., 'Conversational Agent UI', 'Document Intelligence Console') including golden-path templates, scorecards, and documentation, aligned to internal developer-platform (IDP) practices.
• Partner with platform/infra teams to expose infrastructure and model-routing configuration (e.g., cost-aware model selection between Gemini Flash/Pro) through simple, safe self-service UI controls rather than backend-only configuration.
• Establish frontend observability (RUM, tracing of AI response latency, token-stream monitoring) integrated with platform-wide SRE/observability tooling leveraging Dynatrace SaaS.
• Build production-grade full-stack features spanning UI, BFF (Backend-for-Frontend) layer, and orchestration services that call LLM/agent backends (Vertex AI Gemini, Google ADK multi-agent workflows).
• Implement secure, low-latency streaming pipelines (Server-Sent Events / WebSockets) for token-by-token AI responses, agent 'thinking' states, and tool-execution progress indicators.
• Integrate authentication/authorization (OAuth2 client credentials, API Gateway via Apigee X) into frontend session and API-call flows, ensuring B2B and B2E security compliance.
• Collaborate with backend/platform engineers on API contracts, chunking/retrieval metadata exposure (for citation rendering), guardrail/response-safety signal surfacing, and error/fallback UX for model or agent failures.
• Evaluate and select appropriate technologies and tools to support the development and deployment of products on the eCommerce foundation layer
• Collaborate with stakeholders to understand business needs and requirements, and translate them into technical and non-functional specifications
• Work with Product teams to understand their pain points and increase the Developer Experience through Platform Engineering Capabilities
• Strategize work with senior leaders across Ford's Enterprise Architecture, IT Operations to make significant, measurable impact on the eCommerce Platform
• Experience in Cloud FinOps to optimize Cloud Infrastructure Consumption Cost

Responsibilities

• AI-Native Chat Agent UI: Production conversational interface with streaming responses, multi-agent trace visualization, citation/grounding display, and tool-call transparency, built with React 18+/Next.js (App Router, Server Components/Actions).
• AI SDK Integration Layer: Frontend integration React.js-Node.js (Typescript) / LangChain.js / Google ADK web clients for streaming completions, function-calling UI, and structured output rendering.
• Design System Component Library: Versioned, themeable component library (Storybook-documented) with AI-specific components - chat bubbles, agent step timelines, feedback thumbs-up/down + RLHF capture widgets, prompt inspector, retrieval-source cards.
• Real-Time Streaming UX: SSE/WebSocket-based streaming UI patterns, optimistic UI, token-level rendering, and graceful degradation/fallback states for model latency or failure.
• State Management: Robust conversation/session state handling using modern state libraries supporting multi-turn context and agent memory visualization.
• Accessibility Performance: WCAG 2.1 AA compliant, Core Web Vitals-optimized AI UI, with performance budgets for time-to-first token and time-to-interactive.
• Testing Suite: Component (Jest/RTL), E2E (Playwright) test suites including deterministic mocking strategies for non-deterministic LLM outputs.
• Developer Self-Service Portal UI: Internal web console for provisioning AI resources, viewing agent/RAG pipeline health, and managing environment configs, aligned to Cloud Platform self-service provisioning.
• BFF / API Layer: Backend-for-Frontend service (FastAPI / Node.js) mediating between UI and AI backends, handling auth, request shaping, response streaming, and caching.
• Agent RAG Orchestration Integration: Full-stack wiring to Google ADK multi-agent workflows (intent classification, routing, guardrails) and Vertex AI RAG Engine / Gemini for grounded, cited retrieval-augmented responses.
• Vector Search Retrieval UI-to-Backend Contract: End-to-end pipeline exposing embeddings/vector search results (BigQuery, Vertex Vector Search) with chunk-level metadata surfaced in the UI for source citation.
• Secure API Gateway Integration: Full-stack OAuth2 (client-credentials) authentication flows through Apigee X / API Gateway, including token refresh, session handling, and role-based UI rendering.
• Cost-Aware Model Routing UI: Full-stack feature exposing dynamic LLM routing (e.g., Gemini Flash vs. Pro) based on query complexity, with a UI toggle/telemetry dashboard for cost and latency visibility.
• Deployment on Kubernetes (GKE): Containerized frontend and BFF services with CI/CD pipelines (GitHub Actions/Cloud Build), deployed via GKE with progressive delivery (canary/blue-green) and environment promotion aligned to platform standards.
• Observability Feedback Loop: Full-stack telemetry capturing latency, token usage, guardrail triggers, and user feedback, feeding back into model/prompt evaluation dashboards.
• Reference Architecture Golden Path Repo: A reusable, documented starter template (frontend + BFF + agent integration) published to the internal platform catalog for rapid onboarding of new AI product teams.
• Platform-as-a-Product mindset: Define and track adoption, developer satisfaction (DevEx/NPS), and time-to-first-deploy metrics for teams consuming the AI UI SDK/component library.
• Service Catalog ownership: Publish and maintain frontend 'golden path' templates as catalog entries (e.g., via Backstage-style internal developer portal) with clear SLAs, ownership, and versioning.
• Self-Service Enablement: Extend self-service cloud infrastructure provisioning into the presentation layer - enabling product teams to self-provision chat UIs, agent consoles, and dashboards without platform-team tickets.