Chitimella Praneeth
[email protected] 7032743065 Chitimella Praneeth PransCh
Education
Bachelor of Computer Science and Engineering (AI and ML), 12/2021 – present
Keshav Memorial Institute of Technology Hyderabad
CGPA: 8.86
Intermediate in MPC, Tapasya Junior College 07/2019 – 05/2021
Percentage: 96.4% Hyderabad
Work Experience
Teaching Assistant, International Finishing School KMIT 04/2022 – 06/2022
Facilitated individualized Python instructions for 50+ US clients, achieving a 10%
• Hyderabad
average improvement in file handling and OOPs concepts.
Troubleshot, resolved complex code issues, ensuring smooth python development;
•
reduced bug reports by 50% and improved overall code quality.
Projects
Medical Chat Bot using Local CPU, Source code
Developed a sophisticated medical bot based on a large learning model utilizing the quantized LLama-2 7b
•
model on a local CPU.
Implemented Sentence Transformers for sentence embeddings and Faiss CPU for vector storage of
•
embeddings.
Incorporated a conversational interface using the Chainlit library.
•
Aerial images to Map Images Generator, Source code
Directed the creation of a Pix2Pix GAN model, trained on diverse set of 50,000 satellite and aerial images.
•
Achieved a 82% accuracy in generating maps from aerial data, boosting the reliability of urban planning
•
initiatives.
The generator architecture has been modified to a new design known as PatchGAN, incorporating a 10%
•
adjustment in learning rate.
Leveraged an advanced tech stack, employing PyTorch, Keras, and TensorFlow frameworks, alongside
•
autoencoders, to optimize the efficiency of the project.
Object Detection using YOLO, Source code
Utilized YOLOv5 computer vision pipeline for object detection and localization within images.
•
Trained on COCO dataset which contains over 330,000 images, each annotated with 80 object categories and 5
•
captions describing the scene.
Web App performs object detection using the YOLOv5 model, draws bounding boxes around detected objects,
•
and returns the processed image with annotations
Multi-Modal Organ Segmentation and Diagnosis
OrgSeg is a web app that utilizes UNET transformers for precise segmentation of brain tumors, lungs X-rays,
•
abdominal CT scans.
Integrated a GenAI medical bot powered by LLaMa-3 vector embeddings into OrgSeg, enabling intelligent
•
recommendations for diagnosis and treatment planning.
Technical Skills
C/C++/Python/Java HTML/CSS/JavaScript Git/Github Data Structures and Algorithms
React.js/Node.js PyTorch/TensorFlow Object Oriented Programming Computer Vision Keras
Publications
EXPLORING THE INTERSECTION OF MACHINE LEARNING AND IMAGE PROCESSING, IIPV3EBS11 series
Harnesses CNNs and deep neural networks for precise season prediction.
•
Employs advanced image processing techniques to analyze and interpret data.
•
Integrates seamlessly with a weather API for enhanced weather forecasting capabilities.
•
