B.Tech student in Electronics and Communication Engineering at IIIT Dharwad (2023–2027), pursuing a minor in Generative AI and Large Language Models. I build production-oriented systems across three areas: edge AI deployment, LLM-based products, and data engineering.
- 🔭 Currently building ListenMe — an always-on, local, real-time audio transcription app with RAG-based long-term memory
- 🧠 Working across RAG pipelines, fine-tuning, model editing, and agentic workflows
- ⚡ Deploying models to edge hardware (NVIDIA Jetson Orin Nano) with TensorRT/ONNX
- 🛠️ Prior internship: geospatial data pipelines at the School of Planning and Architecture, Bhopal
MLops-Network-Security End-to-end MLOps pipeline for phishing URL detection. FastAPI · MLflow · DagsHub · AWS S3 · Docker · ECR/EC2.
Personal RAG Assistant — live at chat.nirbhay.engineer Deployed retrieval-augmented generation assistant.
ListenMe Always-on local real-time audio transcription app using NVIDIA NeMo/Nemotron ASR, with ChromaDB for long-term memory, SQLite for raw transcripts, and a fine-tuned Gemma model for end-of-day summarization.
RoastMate QLoRA fine-tuning of Gemma-2 2B with an integrated RAG pipeline; deployed via Gradio on Lightning AI.
ROME on GPT-2 Implementation of Rank-One Model Editing for direct factual knowledge edits in a transformer.
Languages: Python, Bash, SQL ML/AI: PyTorch, TensorFlow, Hugging Face, LoRA/QLoRA fine-tuning, RAG (ChromaDB, Pinecone) Edge & Deployment: NVIDIA Jetson Orin Nano, TensorRT, ONNX, Docker Infra & Data: FastAPI, MLflow, AWS (S3, EC2, ECR), n8n, MySQL Other: Git, Linux, computer networking (IP, DHCP, NAT, BGP, SDN)
nirbhaygupta.me · nirbhaygupta4113@gmail.com