PSGPrem Sai Gadwal
Back to Portfolio

Work Experience

A detailed timeline of my professional journey, roles, and key contributions at each company.

Publicis Sapient
IndiaJune 2024 - Present

Roles & Projects

AI Studio

Led frontend development for an enterprise-grade AI multimedia studio processing 10,000+ image generation requests daily using Stable Diffusion, Flux, and Imagen 3.

Achievements:
  • Led frontend development for an enterprise-grade AI multimedia studio processing 10,000+ image generation requests daily using Stable Diffusion, Flux, and Imagen 3
  • Developed 'PromptBoost', an internal AI tool that auto-tuned user prompts based on user behavior analytics, improving image generation success rate by 25%
  • Reduced load time by 40% by implementing efficient state management and optimizing API calls for high-volume asset rendering
  • Designed and built a real-time streaming Chat UI with multi-modal capabilities (text, image, audio) that increased user engagement by 35%
  • Implemented custom React hooks to streamline LLM integration, reducing boilerplate code by 50% and accelerating feature development
  • Established performance benchmarks and implemented monitoring for AI systems, maintaining 99.9% uptime and sub-200ms response times
Technologies:
Next.js
TypeScript
React
Stable Diffusion
Flux
Imagen
LLM Integration
WebSockets
Engage Next

Scaled Quiz Next platform to handle 5,000+ concurrent users with real-time leaderboards and interactive elements.

Achievements:
  • Scaled Quiz Next platform to handle 5,000+ concurrent users with real-time leaderboards and interactive elements
  • Achieved 98% test coverage while integrating new AI features, resulting in zero critical bugs during production launch
  • Engineered multilingual virtual assistant with context-aware responses using vector databases and RAG, supporting 8 languages
  • Implemented real-time data synchronization using WebSockets, reducing latency by 60% for competitive quiz features
  • Optimized image generation pipeline, cutting processing time from 5s to 1.2s while maintaining high-quality outputs
Technologies:
Next.js
TypeScript
Python
FastAPI
WebSockets
RAG
Vector Databases

Key Skills Applied (Overall)

Next.js
TypeScript
Python
FastAPI
Stable Diffusion
Flux
Imagen
LLM Integration
RAG
WebSockets
React
Node.js
LangChain
Publicis Sapient
IndiaJune 2022 - June 2024

Roles & Projects

Quiz Next

Developed end-to-end AI-powered quiz generation system using RAG to automatically create quizzes from transcripts and session content.

Achievements:
  • Developed end-to-end AI-powered quiz generation system using RAG to automatically create quizzes from transcripts and session content
  • Built scalable backend architecture on GCP that processed 500+ concurrent quiz sessions with real-time scoring and analytics
  • Reduced question generation latency by 65% through caching strategies and optimized model inference
  • Created dashboard with real-time analytics that increased client adoption of the platform by 45%
Technologies:
Next.js
TypeScript
Python
FastAPI
GCP
RAG
LLM Integration
Codebuddy - VS Code Extension

Co-developed VS Code extension for AI-assisted coding that achieved 3,000+ installs within first month.

Achievements:
  • Co-developed VS Code extension for AI-assisted coding that achieved 3,000+ installs within first month
  • Architected integration layer for multiple LLM providers, allowing seamless switching between models
  • Implemented token-efficient code chunking algorithm that reduced API costs by 30% while improving code context understanding
  • Built streaming response UI with syntax highlighting and code completion that reduced development time for users by 20%
Technologies:
VS Code Extension
TypeScript
LLM Integration
Node.js

Key Skills Applied (Overall)

Next.js
TypeScript
Python
FastAPI
RAG
LLM Integration
VS Code Extension
GCP
LangChain