LLMs & RAG
Advanced
Hello, I'm
Keshav Garg
AI Engineer
Hello, I'm
AI Engineer
1+ Years
AI Engineering
B.Tech Bachelors Degree
(2021-2025)
Hi, I'm Keshav Garg, a passionate AI Engineer with over a year of professional experience in building enterprise-grade AI solutions. Currently working at Kroolo, I specialize in Multi-Agent Systems, Large Language Models (LLMs), and Retrieval-Augmented Generation (RAG) architectures. I hold a B.Tech in Computer Science from Amity University, Noida, and have successfully delivered innovative AI projects that have transformed business operations.
Explore My
• Revamped Chat with Anything module using advanced text extraction algorithms, reducing processing latency by 50%
• Rebuilt AI agent ecosystem with 250+ integrated tools, transforming prompt-engineered agents into sophisticated agentic systems
• Led technical design reviews and managed full project lifecycle from requirements to production deployment
• Engineered multimodal Deep Research Multi-Agent System with 97% accuracy for fact-checking and misinformation detection
• Developed NLP2SQL system using RAG architecture, achieving 98% query accuracy with automated chart generation
• Built YOLO-based Object Detection Model with custom fine-tuning, eliminating 100% manual CCTV monitoring
• Implemented a Retrieval-Augmented Generation (RAG) pipeline, significantly improving data retrieval accuracy
• Created automated dashboard using user behavior data, eliminating manual data entry and reducing processing time
• Collaborated with cross-functional teams to deliver scalable, high-impact solutions
Advanced
Advanced
Experienced
Experienced
Intermediate
Advanced
Browse My Recent
Human-like AI WhatsApp assistant with dual-memory architecture using PostgreSQL and Pinecone for real-time, context-aware conversations with multimodal inputs.
Autonomous AI support agent with web scraping pipeline using Selenium to extract help articles, featuring LLM-powered search, generation, and verification tools.
Legal document analysis platform using RAPTOR for hierarchical document understanding and automated PDF summarization with interactive "Chat with PDF" functionality.
Video colorization pipeline using deep learning models for automatic colorization of black and white videos, featuring advanced neural network architectures and optimized processing workflows.
Get in Touch
