Chandrakanth Boppudi

● Led a 10-member team, managing project deliverables with a 100% on-time completion rate and no data quality issues. ● Developed data pipelines to train a leading LLM, including tasks such as data cleaning, exploratory data analysis (EDA), feature engineering, machine learning model development, and model evaluation. ● Delivered over 2500 Colab notebooks to the Gemini training team at Google, incorporating data visualizations such as choropleths, waterfalls, word clouds using Python, Numpy, Pandas, Matplotlib, Seaborn, and Plotly, exceeding the original project scope by 25%. ● Created fine-tuning data using SFT and RLHF, including 500+ unique API call scenarios using real and synthetic data. ● Designed and executed comprehensive prompt testing for AI model performance, identifying and resolving over 10 critical issues across areas such as supervised & unsupervised ML, SQL, statistical & data analytics, resulting in a 20% increase in model responsiveness and accuracy.

● Developed AI skill development content (Python, SQL, ML, DL, NLP, OpenCV), benefiting over 50 team members. ● Earned top-tier feedback from 95% of attendees for clear communication and hands-on mentoring in data science concepts.

● Trained, tested, and deployed ML/DL models using scikit-learn, TensorFlow, PyTorch. Cut preprocessing time by 40% using statistical analysis and regex. Extracted key entities (names, phone numbers) with SpaCy and regex. ● Led NLP initiatives for Tata Communications and Hubbell projects, improving data classification accuracy to 95%+ and field extraction accuracy to 98% by fine-tuning SVC models and deploying Azure AI workflows.