Chandana Sasumana

Mattam Street, 534260, Palakol, India.

Work

Qwipo

|

Machine Learning Intern

Summary

Develop and integrate image processing solutions and APIs to improve system efficiency and experience

Highlights

Integrated image processing into a Flask using OpenCV for background removal and PIL for resizing with endpoints

Developed a model for SKU similarity using cosine similarity and TF-IDF on 50k SKUs, achieving 93% accuracy

Developed an API for ONDC application using BecknProtocol, ensuring integration with Node.js and Express

Optimized CockroachDB SQL queries and integrated HTML/TypeScript for improved data retrieval efficiency

Key Skills: OpenCV, PIL, Flask, Node.js, Express, CockroachDB SQL, HTML, TypeScript, TF-IDF

Education

Indian Institute of Technology, Kharagpur

Btech Civil Engineering, Micro Specialization - Artificial Intelligence Grade: 7.72 CGPA

Awards

Selected for AmazonML Summer School

a program focused on machine learning topics led by Amazon scientists

Earned Oracle Cloud Infrastructure 2024 Generative AI Professional certification (85%)

showcasing expertise in LLM architecture, OCI Generative AI Services, and AI application development with RAG and LangChain

Skills

Languages

C, C++, Python, MySQL, Matlab.

Fundamentals

DSA, Image Processing, Statistics, ML, DL, NLP, GenAI.

Libraries

OpenCV, NLTK, PyTorch, Sklearn, Pandas, Transformers, Seaborn, Matplotlib, NumPy.

Models

Linear Regression, CNN, VGG19, RNN, LSTM, GRU, BERT, Gradient Boosting, LLama2.

Big Data Technologies

FAISS, ChromaDB.

Tools

PyCharm, Docker, Kubernetes, Google Colab, Jupyter.

Projects

Credit Card Fraud Detection

Summary

Build an explainable fraud detection system to identify credit card fraud while maintaining transparency

SQL Query Retrieval

Summary

Enhance accessibility by converting SQL queries into natural language responses for improved interaction

Sentiment Analysis

Summary

Build a sentiment analysis model using a fine-tuned BERT to classify sentiment in Twitter data