Supreetha Naik

• Total 5+ years of experience with 3+ years as a Data analyst. Strong skills in SQL, building reporting/KPI dashboards using Tableau and power BI.
• Experience in building data models, ETL processing on AWS, with proficiency in SQL, NoSQL databases; experience working with Databricks, Hadoop and Spark.
• Immense knowledge and experience in computer vision and NLP, and GPT3 models.
• Built Machine Learning models using Python libraries, Pandas, Skicit-learn, Tensorflow, and PyTorch.
• Agile project lead, technical project management, and cognizance in Agile requirement design.

• Ship detection and masking using custom CNN (Unet) and pre-trained models (Resnet, Mask R CNN, Mobile Net).
• Google competition of American Sign Language (ASL) recognition and conversion to text using LSTM and Transformer.

 STACK OVERFLOW RANKING SYSTEM USING NLP 

• Built stack overflow search algorithm using BERT transformer model with LSTM, ELMO.
• Conducting research on Facebook AI similarity search for building ranking model.

 MUSIC RECOMMENDATION SYSTEM FOR MENTAL HEALTH 

• Built music recommendations using ML models – XGBoost, Random Forest, KNN, SVM. Utilized SMOTE for data balancing and developed a user-interactive chatbot for the recommendation system.
• Implemented collaborative filtering using matrix factorization using SVD to enhance the system.

 ENERGY CONSUMPTION ANALYSIS USING BIG DATA TECHNOLOGIES 

• Developed ETL pipelines to extract, transform, load and integrate 10GB of data from various sources into data warehouse Redshift from S3; utilized glue crawlers for processing

 DECENTRALIZED SMART GRID LOAD PREDICTION USING ML 

• Smart grid stability prediction using machine learning models leveraging SVM, Random Forest,XGBoost to classify load performance.
• Implemented hyperparameter tuning to optimize machine learning model performance, increasing accuracy by up to 5%.