About
I'm Kandimalla Bruhadev — an AI Automation Engineer and published researcher building intelligent systems.
AI/ML Engineer specializing in Large Language Models, RAG, and AI-driven application development. I build intelligent systems that solve real problems — from RAG-based legal platforms to AI-powered HRMS and financial intelligence tools.
Published researcher — my IEEE paper on JurisGPT, a citation-grounded legal assistant for Indian startup and corporate law, covers a 47,867-document corpus across ten statutes, hybrid retrieval hitting Recall@5 of 0.84 and MRR of 0.95, and answers judged 80% citation-faithful with a 2.5% severe-hallucination rate.
With deep expertise in LLMs, RAG pipelines, and multi-agent systems, I combine AI-powered development tools with strong engineering fundamentals to deliver products that are fast, beautiful, and production-ready.
What I do
Vibe Coding
AI-first development using Claude Code, Cursor AI, v0, Bolt & Windsurf to ship at 10x speed
Claude Code Expert
Deep expertise in Claude for code generation, debugging, refactoring, and system architecture
Prompt Engineering
System prompts, few-shot learning, chain-of-thought, and tool use patterns
RAG Pipelines
Semantic chunking, embeddings, vector databases, hybrid search, and reranking
Agent Development
Multi-agent systems with CrewAI, tool calling, memory management
AI Tools Mastery
Copilot, Lovable, Gemini, ChatGPT, Windsurf — fluent in every AI coding tool
LLM Evaluation
Custom scoring frameworks, hallucination detection, safety checks
Full Stack Dev
Next.js, React, Python, FastAPI, databases, and cloud deployment
Published Research
IEEE paper on JurisGPT: citation-grounded legal RAG, 47.8K-document corpus, benchmarked with significance testing