Swetha Gopikumar Sreeja

- Collaborated closely with the Office for the Arts on the Educational Lending Rights (ELR) project, a government initiative.
- Conducted meticulous analysis and processing of data gathered from schools across Australia and New Zealand, optimizing efficiency by 30%.
- Integrated ELR data seamlessly with the Schools Catalogue Information Service (SCIS) using Python, SQL, and data engineering skills.
- Designed and implemented PowerBI reports, creating informative dashboards for the marketing team.
- Automated processes for mailing lists and subscription cancellations using Power Automate, reducing manual efforts by 78% and significantly streamlining SCIS workflow.
- Tracked project targets for every month, contributing to the overall success of the ELR project. Prepared detailed reports for educational ministers, providing crucial insights and acting as a statistician on ELR matters.

- Worked on a significant research project at Melbourne Data Analytics Platform (MDAP) aimed at "Understanding Evolution and Immunity of Amphibian Genomes" with the goal of preventing the extinction of a vulnerable amphibian species.
- Applied advanced genome assembly methods including RagTag and Abyss to analyze amphibian genomes, with a particular focus on those exceeding 1 terabyte in size.
- Carried out a comprehensive comparative analysis by evaluating the accuracy of genome assembly methods using BUSCO scores, contributing valuable insights into the optimization of genomic data processing.
- Enhanced the deep learning model, terrier (Transposable Element Repeat Result classifIER), by incorporating sophisticated techniques such as focal loss, spectral clustering, and PCA, using Tensorflow and Keras.
- Achieved a remarkable classification accuracy of 93.87% in identifying large genome repeats through the integration of these techniques.
- Utilized the high-performance computing capabilities of SPARTAN at the University of Melbourne to execute the computationally intensive tasks required for the deep learning model, showcasing technical expertise in high performance computing(HPC), bioinformatics and computational biology.

- Employed deep learning techniques to predict profits and potential defects in machines supplied every month at Caterpillar, enhancing accuracy and precision, using Tensorflow and Keras, resulting in a 20% increase in profits.
- Generated comprehensive PowerBI dashboards and reports for stakeholders, providing insights into product performance, projected profits, and current targets.
- Streamlined the markup process for design engineers through the implementation of ML algorithms, including decision trees, deep CNNs, contributing to increased team efficiency by 25%.
- Implemented a job scheduler for the design team, incorporating ML techniques and data warehousing concepts to streamline workflows and enhance overall productivity, resulting in a 82% reduction in processing time.

- Automated the scheduled updating of sales models, using REST and SOAP services, whenever any of the sales models advance to new model specific changes, using C#.
- Developed a multi-dashboard website using Angular, SQL and Python, giving more attention to user experience and reusability.

- Incorporated a feature using C# and PowerShell to automate batch processing of files across diverse libraries, significantly enhancing processing speed for businesses by 60%.
- Crafted a bespoke PowerShell drive from scratch, leveraging a hierarchical structure derived from XML files. Integrated seamlessly with MongoDB to establish the files and folders structure, aiming to enhance search efficiency and streamline data retrieval processes.
- Created an interactive Windows application that customizes the plug-ins and generalizes templates for dashboards, thereby automating UI creation and necessary reporting tools using WPF/C# in the .NET framework.
- Devised rule specific actions to support customizations for plug-ins and for the reporting tools to send out notifications/alerts (actions) based on these rules.
- Integrated with MongoDB and SQL databases, strictly following MVVM architecture.
- Leveraged template creation by implementing a hierarchical search in the XML files.