Chenji Ipshitha Royal

Passionate about pushing the boundaries of machine learning algorithms through hands-on experimentation and implementation within Data Science teams. Experienced in deploying and fine-tuning ML and DL models to solve complex problems. Proficient in generative AI techniques, specializing in crafting chatbots through prompting skills and leveraging LLMs. Skilled in experimenting with various frameworks such as LangChain to optimize performance and drive innovation in natural language processing tasks. Eager to showcase my developed skills by creating new products/projects and engaging in projects that further enhance my expertise and learning journey.

• Personalized College and Career Chatbot
  March 2024 - Present
  Currently developing a sophisticated personalized college and career chatbot, designed to streamline the selection process for students by utilizing Retrieval-Augmented Generation (RAG) technology. The chatbot integrates extensive data from various sources, including college information, program details across different branches, and external career-related APIs like ONET, which provide in-depth career data. Embedded data from these sources is used to generate accurate, contextually relevant responses tailored to individual student preferences. The main challenge lies in addressing issues like hallucinations, maintaining accuracy, and ensuring high-quality responses from the LLM. This project exemplifies the fusion of cutting-edge AI technology with education, offering students a personalized, informed pathway to selecting colleges and careers while optimizing the application process.
• Credit Risk Management System
  Jan, 2024 - April, 2024
  Developed a credit risk management system leveraging machine learning algorithms such as Random Forest and XGBoost to predict default probability. The project utilized extensive financial data, including credit scores, transaction history, and customer demographics. Key metrics like AUC-ROC, precision-recall, and accuracy were used to evaluate model performance, ensuring high prediction accuracy and minimizing false positives. This project improved decision-making for loan approvals and risk mitigation in financial institutions.
• Student tutoring Application
  June, 2023 - Jan, 2024
  Led the development of a groundbreaking tutoring app powered by Generative AI, integrating LLM GPT models and prompt engineering to tailor conversation flows. Implemented innovative features such as multimedia support for image and video content, streaming for interactive sessions, and a custom code interpreter for dynamic math problem solving, resulting in enhanced student engagement and learning efficacy.
• Charting Application using pandasai
  April, 2023 - June, 2023
  Basic charting generator, which uses data convert to pandas dataframe and plot fascinating charts/reports using various visualisation packages. Additionally, implemented logic to accept natural language user input to give result in natural language with charts.
• Image classification
  August, 2022 - May, 2023
  Developed an image classification model using Convolutional Neural Networks (CNN) for identifying objects in a dataset of 50,000 labeled images. Utilized TensorFlow and OpenCV for pre-processing and model development. Implemented data augmentation techniques to enhance model performance. Achieved 92% accuracy, optimizing with Adam optimizer and using categorical cross-entropy as the loss function. Evaluated model performance with metrics like precision, recall, and F1-score, ensuring robust classification across multiple categories.

• 2020-04 Statistics & Mathematics for Data Science & Data Analytics by Udemy
• 2020-05 Machine Learning using Python - Course offered by IBM
• 2020-10 Basics of AI and ML from Future skills NASSCOM
• 2021-04 Cloud Computing Basics from Coursera
• 2022-06 Python for Data Science and Machine Learning
• 2022-08 NLP in Python by Udemy
• 2022-11 NLP With Transformers in Python by Udemy
• 2023-05 Prompt Engineering from LinkedIn Learning
• 2023-07 Langchain- DevelopLLM powered applications with Langchain by Udemy
• 2024-01 Accurate Prompting with Anthropic’s Claude by LinkedIn Learning
• 2024-02 Advanced Prompting Techniques by LinkedIn Learning
• 2024-03 Meta Llama 2 Foundational Model with Prompt Flow by LinkedIn Learning

• Follow Blogs - Medium.com, Towards data Science, TLDR NewsLetter
• E-Mail: ipshithachenji@gmail.com
• LinkedIn: https://www.linkedin.com/in/ipshitha-chenji
• GITHUB: https://github.com/Ipshitha13