Skip to content
View Deepthit-23's full-sized avatar
0 followers · 1 following

Achievements

Achievement: YOLO

Achievements

Achievement: YOLO

Block or report Deepthit-23

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.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. 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
Deepthit-23/README.md

Hi, I am Deepthi 👋 I am a B.E. student in Artificial Intelligence and Machine Learning at BMSIT&M, Bangalore. I love building things that actually work, from training ML models to deploying backend systems that serve real users.

What I work with

Languages: Python, C, C++ Backend: FastAPI, Flask, REST APIs ML/AI: TensorFlow, Keras, Transfer Learning, NLP Databases: SQL, MySQL, MongoDB Tools: Git, Google Colab, Hugging Face, Vercel, GCP Game/XR: Unity, XR Interaction Toolkit, Meta Quest 2/3

Projects 🩺 WoundScan AI An AI powered wound classification web app that photographs a wound and instantly returns the wound type along with first aid instructions across 10 categories. Built a MobileNetV2 transfer learning model trained on 2,940 images with a FastAPI backend deployed on Hugging Face Spaces and a React/Vite frontend on Vercel.

🛡️ Lightweight Real-Time LLM Safety Gateway A high performance reverse proxy safety gateway that enforces security guardrails for LLMs with sub 25ms CPU based latency. Uses a multi layered defense pipeline combining regex filtering, heuristic analysis and MiniLM-L6-v2 semantic embeddings to detect malicious prompts. Includes a real time analytics dashboard built with Streamlit.

🗣️ VocaLing A voice based AI language learning application built with Flask that enables real time spoken conversation practice. Integrated LLaMA 3.1 via GroqCloud for context aware responses, transformer based NLP for intent recognition and Text-to-Speech for natural audio feedback. Designed with a modular RESTful backend for scalability.

🥽 VR Lab Safety Trainer An immersive VR training application for Meta Quest 2/3 built in Unity that teaches laboratory safety through four hands-on scenarios — PPE usage, fire extinguisher operation, chemical hazard identification and chemical mixing safety. Features a scoring system, progress tracking and an intuitive VR menu. Built using Unity 2022.3 LTS and the XR Interaction Toolkit.

A little more about me

I practice DSA on LeetCode to keep my problem solving sharp I enjoy AR/VR projects to keep myself from getting too comfortable in one domain I am a general learner at heart and always exploring something new

Pinned Loading

  1. LLM-safety LLM-safety Public