Datha
Hyderabad, Telangana, India
Experience: 0 years
Open to: Full-Time
Education: Masters
Availability: Within 30 Days
Skills: Azure, Jupyter Notebooks, LLMs/ChatGPT, Matplotlib, No Code, NumPy, pandas, Pentaho, Power BI, Problem Solving, Python3, scikit-learn, TensorFlow
Previously worked at: IT Service Companies
Assessment Score: 100
- Professional Experience and Expertise: With over 4 years of experience in data science and analytics, currently serving as a Functional Analyst at HDFC Bank, specializing in automating financial processes, predictive modeling, and risk forecasting. Previous experience as a Data Scientist at Tata Consultancy Services, focusing on customer churn modeling and sentiment analysis.
- Educational Background: Holds a Post Graduate Diploma in Big Data Analytics from Goa Institute of Management (expected 2024) and a Post Graduate Diploma in Full Stack Development from IIIT Bangalore (2020). Completed a B.Tech in Information Technology from SRM Institute of Science and Technology, Chennai.
- Machine Learning and AI Proficiency: Expertise in machine learning techniques including supervised, unsupervised, and reinforcement learning, deep learning (RNN, CNN, GAN, VAE), and text analytics. Skilled in using tools like TensorFlow, Keras, Scikit-Learn, NLTK, and Hugging Face for AI model development and deployment.
- Project Leadership: Led impactful projects such as the development of an AI-powered admissions chatbot utilizing FLAN-T5 and Sentence-BERT, and research into optimizing diffusion models for enhanced image synthesis, achieving significant improvements in computational efficiency and image quality.
- Achievements in Automation and Risk Management: Successfully automated salary identification for over 7 million transactions at HDFC Bank, reducing manual intervention by 45%. Improved loan sanctioning accuracy and churn prediction, contributing to a 15% reduction in churn for car finance products.
- Academic and Research Excellence: Recognized for academic achievements, including a top 5% finish in GIM’s BITATHON Data Analytics Championship and research excellence in diffusion models, enhancing image reconstruction techniques with 30% lower computational costs and 15% better image quality.