Komali Devi Valluru

Experienced data engineer with hands-on experience in designing, building, and maintaining scalable and reliable data pipelines using technologies such as ELT, SQL, and Spark. Skilled in collaborating with cross-functional teams to deliver successful data-driven solutions.

• Azure Fundamentals – Microsoft
• Infosys Global Agile Certification – Infosys
• Data Science Foundations Level1 & Level 2 – IBM
• Python for Data Science – IBM

Animal Breed Classifier:

• The project involved developing an animal breed classifier that uses Convolutional Neural Networks (CNN) to identify the breed of an animal from input images.
• Two approaches were used for model building: A classification model from scratch , and a model using Transfer Learning with pre-trained models, VGG16 and InceptionV3. The models were compared in terms of accuracy and efficiency.
• The project aimed to investigate the effectiveness of CNN architecture and Transfer Learning for image classification tasks.

3D Expression Recognition:

• The dataset consists of .BND files which have co-ordinates of the expressions on 3D axis. The dataset is processed and transformed as rotated and translated to create new datasets.
• The transformed datasets are compared with the original dataset and models are built with three classification algorithms(Random Forest, Decision Tree, SVM) to analyze the models and how they run on each dataset.

Partner Business Modeling:

• Given a dataset of aggregated rides on a busy Saturday, the project involved analyzing potential bonus options in terms of cost that could incentivize more drivers to go online and help manage surge during weekends.
• The objective is to identify any patterns or correlations among the variables within the dataset.