Sankhamala Palsankhamala

Scientific Assistant at Fraunhofer ISE, Freiburg, Germany

Developed a Graph Neural Network-based classification model to detect diverse errors in power grid models, improving robustness against noisy and missing measurements. Extended Python scripts for data ingestion and published values for various data sources. Involved in implementing new features, bug fixes, code reviews, and integration of REST APIs along with swagger.io code generation.

Master Thesis Student at Fraunhofer ISE

Developed AI-based validation models for power grid models using graph neural networks. Focused on improving robustness to noisy and missing data to enhance real-world applicability of the models.

Research Assistant at Albert Ludwig University of Freiburg

Worked on interactive, context-aware dietary food logging application using Flutter frameworks. Enhanced user experience by easing data entry by 25% through new feature additions. Developed SAP Self Service Portal features for file processing and SAP Automation tools that reduced effort by 20%.

Application Development Associate / Graduate Intern at Accenture Solutions Private Limited, India

Designed and tested a dietary food logging application, improving the data entry process. Worked on SAP portals for improved automation and file processing. Achieved significant performance and automation efficiency gains on portal modules.

Software Developer at Accenture and Fraunhofer ISE

More than 3 years of experience in Java development focusing on backend frameworks like Spring, JPA, and Hibernate, and frontend technologies such as JavaScript, HTML, CSS, and PHP. Experience involves using REST APIs, database management with SQL, MySQL, and SQLite, and cloud deployment on Azure and AWS.

Scientific Assistant at Fraunhofer ISE / University of Freiburg Intelligent Embedded Systems Lab

Developed a Graph Neural Network-based classification model to detect diverse errors in power grid models, improving robustness against noisy and missing measurements to enable real-world applicability. Modified the model to a regression approach to predict specific error parameters in grid components. Used Python, PyTorch, Torch Geometric, and TensorFlow. Responsible for implementation of new features, fixing bugs, code reviews, writing unit and integration tests, embedding auto-generated swagger.io code, and extending Python scripts for data processing. Worked extensively with Git and Gradle.