Praveen Shet G,
frontend systems & AI integration engineer.
12+ years building scalable React & Next.js platforms — now extending audit and enterprise workflows with agentic AI, from Python-based reasoning layers to production LLM integrations.
From pixel-perfect UI to AI-augmented platforms.
I'm a frontend engineering leader with 12+ years shipping scalable, performant web and mobile platforms. My core strength is React, TypeScript, and modular design systems — with a deep focus on performance optimization, state management, and engineering rigor.
At EY, I've extended that foundation into AI integration engineering: building agents and Python-based AI layers on top of audit platforms, integrating Agent Builder and OpenAI GPT-5.4 to power risk analysis, control analysis, and automated process flow diagram generation.
Outside of client work, I mentor engineers, lead code reviews, and I'm currently building Asset Extract, a solo PropTech platform for property valuation.
Languages, architecture & infrastructure
Languages & Frameworks
Architecture & State
AI & Integration
Tooling, Data & Cloud
Experience
Reverse-chronological — including client engagements delivered through parent consulting firms.
- Built the frontend architecture from scratch using React, Next.js, TypeScript, and Redux for a scalable, high-performance audit platform.
- Architected monorepo setup using Storybook, improving reusability and cross-app consistency by 40%.
- Built AI agents and a Python-based AI layer for the audit process, integrating with Agent Builder and OpenAI GPT-5.4.
- Enabled risk analysis, control analysis, and automated process flow diagram generation for the audit platform.
- Architected cross-platform mobile components in React Native, optimized for larger data sets.
- Managed and mentored a frontend team of 3 engineers for finance & merchant platforms.
- Implemented micro-frontend architecture for loans, accounts, and other banking service modules.
- Improved core web vitals and app responsiveness by 35% through caching, lazy loading, and bundle optimization.
- Contributed to the QuickBooks UI library with reusable components: grids, OTP inputs, modals, loaders.
- Migrated a Drools rule-engine-based application to the latest Java version, working with the RedHat team on JBoss server configuration.
- Delivered Java training to a testing team of 150 members.
- Worked as a Groovy/Grails developer building data-driven web applications with D3.js and C3.js visualizations and analytics.
Side project spotlight
Asset Extract
A self-initiated PropTech platform streamlining property valuation, collateral assessment, and report generation for valuers, banks, NBFCs, and financial institutions.
- Full-stack workflow digitization for property inspection, document collection, and valuation data capture.
- Secure Google OAuth authentication with role-based access control on cloud-hosted infrastructure.
- Geospatial intelligence via Google Maps APIs for site visualization and valuation analysis.
- AI-assisted collateral intelligence supporting risk analysis and valuation decisions.
Leadership & impact
Junior/mid engineers mentored and onboarded, with 3 guided to promotion readiness.
Code reviews conducted, upholding code quality and design consistency.
Engineers trained on Java as part of a testing-team enablement initiative.
Let's build something worth shipping.
Open to conversations about frontend leadership, AI integration, or interesting problems in between.