Akanksha G

I build production-ready Generative AI systems that are low-latency, retrieval-grounded, and enterprise-ready. As a Generative AI Engineer, I specialize in RAG architectures, Agentic workflows, and LLM fine-tuning. My work focuses on reducing hallucinations, improving retrieval accuracy, and designing scalable AI applications from idea to deployment. During my internship, I: • Designed and deployed a RAG-based educational assistant, improving answer relevance by 30% • Engineered AI Agents for quiz generation and multilingual workflows, reducing manual effort by 40% • Implemented LLM response caching and structured retrieval pipelines to reduce API cost and latency I have architected hybrid LLM pipelines (Dataset Matching + SLM + RAG) using LangGraph, fine-tuned TinyLlama using QLoRA, and built privacy-first document intelligence systems for financial data extraction. Core Areas of Focus: • Retrieval-Augmented Generation (RAG) • Agentic AI & Multi-Step Reasoning Workflows • LLM Fine-Tuning (QLoRA, PEFT) • Vector Databases (FAISS, ChromaDB) • LLM Deployment with FastAPI, Streamlit & AWS Tech Stack: Python | LangChain | LangGraph | Hugging Face | OpenAI | Groq | FAISS | ChromaDB | AWS | FastAPI | Streamlit I’m open to opportunities in Generative AI Engineering, Machine Learning Engineering, and AI Systems Development. 📩 Let’s connect if you’re building intelligent AI systems.