Akshayraj Nadar

Driven Data Science Intern ready to thrive in demanding digital intelligence processing environments. Well-informed on latest machine learning advancements. Ready to combine tireless hunger for new skills with desire to exploit cutting-edge data science technology.

PillboxGPT

Personal Project • https://github.com/Akshayrajnadar/PillboxGPT • December 2023- December 2023

• Created a medical chatbot named PillboxGPT by utilizing the open-source Python tool Chainlit to quickly create apps with unique logic and data.
• Included other libraries, including Hugging Face, a community platform for machine learning, llama2 for pre-trained large language models (LLMs) provided by Meta AI, and CTransformer, which uses C/C++ with GGML library for Transformer models.
• Widely utilized the open-source orchestration framework LangChain, which is designed for applications that use huge language models.
• This project improves accessibility by using an intelligent and dynamic conversational interface to inform users about different diseases and their treatments.

Image Classification

Personal Project • https://github.com/Akshayrajnadar/Image-Classifi cation • September 2023 - October 2023

• Developed a CNN-based imagine classification algorithm that can accurately identify 90 different animal classes.
• Shown expertise in both deep learning and web development by designing and implementing an interactive Django interface that lets users submit photos for categorization.
• Utilizing prominent libraries such as TensorFlow, scikit-learn, Matplotlib, PIL, and proficiently handling JSON files.
• This self-initiated project's usefulness is enhanced by the smooth integration of the CNN model with the Django framework, which guarantees an easy-to-use interface.

Diabetes Prediction Model

Personal Project • https://github.com/Akshayrajnadar/Diabetes-Prediction • March 2023 - April 2023

• Developed and deployed a robust neural network model for diabetes prediction using sophisticated deep learning techniques.
• Attained a good 74% accuracy score for the created neural network, exhibiting skill in model training and validation.
• Utilized modern preprocessing approaches to optimize data quality and model performance while curating and cleansing the diabetes dataset.

• The deep learning model is smoothly integrated with the Django framework, ensuring easy communication between the user interface and

the predictive model and therefore improving the overall functioning of this independent project.

• Good team worker
• Skilled in multitasking and problem-solving
• Hard work and enthusiastic