Gohitha Reddy

Analytical and results-oriented Computer Science graduate student with strong expertise in machine learning, data science, and AI system design. Proven experience in building high-performance ML models, conducting data-driven research, and creating real-time predictive dashboards. Seeking data scientist or analyst roles to apply technical depth, and drive impactful insights.

Bi-Directional Caduceus: Large Language Model for Multimodal Diagnosis 

Developed and evaluated a transformer-based clinical reasoning model (Bi-Directional Caduceus) combining medical images, clinical text, and patient vitals. Trained and tested on multimodal datasets with aligned vision-language supervision, implementing attention mechanisms, feature fusion, and comparative baselines.

Facial Emotion Recognition using CNN 

Built a deep learning model on FER2013 dataset to classify facial expressions with 83% validation accuracy. Performed preprocessing, CNN training, and confusion matrix analysis for emotion-based HCI applications.

Geography-Aware Self-Supervised Learning with MoCo-v2 

Implemented a geography-conscious contrastive learning framework using MoCo-v2 on the fMoW remote sensing dataset. Integrated geo-location classification and temporal positives into the contrastive loss, achieving 88.3% accuracy, outperforming standard supervised baselines in classification, detection, and segmentation tasks.

• Google Cloud Professional Engineer
• IBM Data Science Professional
• Microsoft Azure Data Scientist Associate
• TensorFlow Developer
• Deep Learning Specialization