Skip to content
View RohanBabbar's full-sized avatar

Block or report RohanBabbar

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
RohanBabbar/README.md

Hi, I'm Rohan 👋

AI/ML engineer focused on LLM applications, RAG, and model fine-tuning. B.Tech in CSE (AI), Amrita Vishwa Vidyapeetham · Former ML intern at DRDO. Open to AI/ML engineer roles and internships.

🔧 Featured projects

Agentic RAG Router Deterministic agentic RAG built from scratch in Python (no LangChain) that routes queries across different document types.

Mood-Based Story Generator · ▶ Live demo Illustrated story generator with RAG over user-uploaded lore (FAISS + MiniLM), running Llama on Groq with SD-Turbo images, deployed on Modal.

Hinglish Speech Processing Fine-tuned Whisper-small on Hindi FLEURS, cutting WER from 84.2% to 47.5%. Also built a 6-model ROVER-style consensus evaluator and a spell-checker validating 177k Hinglish words.

CWE Vulnerability Classifier Fine-tuned CodeBERT, RoBERTa and CodeT5 to classify software vulnerabilities on the Juliet C/C++ suite, with a leakage-safe split and weighted loss for rare classes.

🧰 Stack

Python · PyTorch · Hugging Face Transformers · PEFT/LoRA · FAISS · Sentence-Transformers · Gradio

📫 Contact

LinkedIn · rohanbabbar913@gmail.com

Pinned Loading

  1. ASR_Research ASR_Research Public

    End-to-end ASR and NLP solutions for conversational Hindi speech processing.

    Python 1

  2. cwe-vulnerability-classifier cwe-vulnerability-classifier Public

    A Machine Learning research project focused on automated Software Vulnerability Classification (CWE) using fine-tuned linguistic models (CodeBERT, RoBERTa, CodeT5) on the Juliet C/C++ Test Suite.

    Jupyter Notebook 2

  3. LLM-Document-Analyzer LLM-Document-Analyzer Public

    A modular Python data pipeline that ingests unstructured text from URLs and files (PDF, TXT), processes it, and extracts structured insights using Google's Gemini LLM.

    Python

  4. mood-story-rag mood-story-rag Public

    Mood Story RAG addresses a critical problem in generative AI storytelling: the inability of generalized LLMs to maintain specific, user-defined "world lore" while strictly adhering to episodic, moo…

    Jupyter Notebook 1

  5. Agentic-RAG-Router Agentic-RAG-Router Public

    Python