Skip to content
View ShathaVarsha's full-sized avatar
1 follower · 1 following

Block or report ShathaVarsha

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
ShathaVarsha/README.md

Hi, I'm Shatha Varsha Sree T 👋

Final-year B.Tech student in Artificial Intelligence & Data Science at Amrita Vishwa Vidyapeetham (B.Tech, 2023–2027). I build AI systems for healthcare and pharma — from EMG-driven pose estimation and multimodal mental-health screening to LLM-and-RDKit-powered drug discovery.

💊 AI & Data Science Intern — Mankind Pharma Limited, May 2026 – Jun 2026

🔬 What I'm working on

  • AI-assisted drug discovery — molecular generation, virtual screening, and toxicity prediction with RDKit + LLMs
  • Multimodal mental-health screening using text, video, and voice biomarkers
  • EMG-based human pose estimation with GPU acceleration and Spark/cloud-scale deployment
  • Quantum machine learning for phenotypic image classification in drug screening
  • Agentic LLMs for neurological speech-disorder diagnosis and voice restoration

🛠️ Skills

ML / AI: PyTorch · Scikit-learn · CNN · LSTM · Feature Engineering · TabPFN · LIME Data Analytics: EDA · Data Cleaning & Preprocessing · Correlation & Lag Analysis · Statistical Analysis · Data Validation Libraries: NumPy · Pandas · OpenCV · FastAPI · RDKit Voice AI: OpenAI · Azure AI Speech Services · ElevenLabs (STT/TTS) Languages: Python · C · MATLAB

📌 Featured Projects

Project Description Tech
AI Agent Molecular Generation for Heart Disease Drug Discovery AI agent pipeline for molecular generation & virtual screening targeting heart-disease drug candidates Python, RDKit, LLMs, SMILES
Multimodal Mental Health Screening (DAIC-WOZ) Multimodal depression & PTSD screening fusing NLP and facial video cues, 88–91% accuracy Python, NLP, OpenCV
EMG-Based Human Pose Estimation (GPU + Cloud) GPU-accelerated CNN-LSTM model for joint-angle prediction from EMG signals, cloud-deployed Python, CNN, LSTM, CUDA, Cloud
Scalable Hand Pose Estimation with Apache Spark Distributed hand-pose estimation from muscle signals, scaled with Apache Spark Python, Apache Spark, CNN/LSTM
Quantum ML for Phenotypic Image Classification Quantum machine learning models for phenotypic image classification in drug-screening pipelines Python, Quantum ML, Image Classification
Agentic LLM for Neurological Speech Disorders Agentic LLM system for diagnosing neurological speech disorders and guiding therapy & voice restoration Python, LLMs, Voice AI

📄 Publications

  • Bio-Inspired Snake Robot for Real-Time Victim Detection in Disaster Environments — Mar 2026
  • Emotion-Aware Image Captioning using Deep Learning — Mar 2026
  • Solar Powered Agricultural Robot for Irrigation (Co-Author) — Feb 2026

🌐 Languages

  • English (Professional)
  • Telugu (Native)
  • Tamil (Basic – spoken & comprehension)

📫 Connect

LinkedIn Email GitHub Stats

Pinned Loading

  1. Scalable-Hand-Pose-Estimation-from-Muscle-Signals-using-apache-spark Scalable-Hand-Pose-Estimation-from-Muscle-Signals-using-apache-spark Public

    Scalable hand pose estimation from EMG muscle signals using Apache Spark. This project processes high-volume biosignal data and predicts hand poses through distributed machine learning pipelines.

    Scala 1

  2. Agentic-LLM-for-Neurological-Speech-Disorder-Diagnosis-Therapy-and-Voice-Restoration Agentic-LLM-for-Neurological-Speech-Disorder-Diagnosis-Therapy-and-Voice-Restoration Public

    An agentic LLM-based system for diagnosing neurological speech disorders and supporting therapy planning and voice restoration. Combines autonomous AI agents with speech and language models to assi…

    Jupyter Notebook 1 1

  3. AI-Agent-Based-Molecular-Generation-and-Screening-for-Heart-Disease-Drug-Discovery AI-Agent-Based-Molecular-Generation-and-Screening-for-Heart-Disease-Drug-Discovery Public

    An AI agent framework for generating and screening candidate molecules relevant to heart disease drug discovery. Automates early-stage compound generation and filtering to accelerate pharmaceutical…

    Jupyter Notebook 1

  4. EMG-Based-Human-Pose-Estimation-Using-GPU-Accelerated-with-Cloud-Deployment- EMG-Based-Human-Pose-Estimation-Using-GPU-Accelerated-with-Cloud-Deployment- Public

    This project implements a real-time EMG-based human pose estimation system. Traditional camera based pose estimation struggles in low-light and privacy-sensitive environments. Instead, we are usin…

    1

  5. Multimodal-Mental-Health-Screening-System-using-Text-and-Video-Cues-DAIC-WOZ-Dataset- Multimodal-Mental-Health-Screening-System-using-Text-and-Video-Cues-DAIC-WOZ-Dataset- Public

    A multimodal machine learning system for mental health screening that detects depression by analyzing both text and video cues from the DAIC-WOZ dataset, combining NLP and computer vision to captur…

    Python 1

  6. Quantum-Machine-Learning-for-Phenotypic-Image-Classification-in-Drug-Screening- Quantum-Machine-Learning-for-Phenotypic-Image-Classification-in-Drug-Screening- Public

    The application of Quantum Machine Learning (QML) to classify drug-induced phenotypic responses in human breast cancer cells, using microscopy images from the BBBC021 dataset. The primary objective…

    Jupyter Notebook 1