Srinath Jayaraman Nagamani Professional profile and links srinath.jayaraman92@gmail.com Education ○ Delft University of Technology Delft, The Netherlands MSc. Computer Science - Data Science and Technology November 2022 ○ B.S. Abdur Rahman University Chennai, India BTech. Electronics and Instrumentation Engineering, First Class May 2014 Competencies ○ Programming Languages: Python, Go, Scala, MSSQL & PL/SQL, COBOL, JCL, IBM DB2. ○ Software, Libraries, Environments, & Tool-kits: Git, Pandas, Apache Spark, Apache Flink, PyCharm, Jupyter Notebooks, Redis, MS Excel. ○ Concepts and Disciplines: - Machine Learning - Clustering (Fuzzy, K-means), classification (KNN, random forest, SVM, decision tree), regression (simple, polynomial, support vector), and deep learning (Convolutional Neural Networks). - Data Engineering - PCA, LDA, & dimensionality reduction, imbalanced data-sets, generating synthetic data (e.g. via SMOTE) & traditional pitfalls of synthetic data. - Data Analysis - Utilising Hadoop & MapReduce for big data processing on AWS, implementing parallel algorithms using the Apache Spark framework, and performing analysis on streaming data via Apache Kafka. ○ Spoken Languages: Fluent in English, Tamil, & Hindi. Previous Experience ○ Cognizant Technology Solutions Ltd. (Chennai, India & Buenos Aires, Argentina) Associate - Support Analyst, Retail July 2014 - May 2019 • Analysed existing code base and solved production bugs. • Automated cumbersome manual tasks that reduced operating time and increased overall system-wide efficiency. • Improved CPU usage by collaborating with the infrastructure and database teams, leading to significant financial savings. • Updated FTP scripts to guarantee adherence to modern security standards. • Spearheaded troubleshooting and cross-reporting between businesses for better performance capture and system wide monitoring. • Frequently collaborated with management to ensure transparent reporting. • Mentored newcomers and fresh recruits. Notable Projects ○ MSc. Thesis: AI-ForestWatch - Reproducing a deep learning algorithm - when unstoppable expecta- tions meet immovable reality - (Python) • Modified an existing deep learning network that analysed high-resolution satellite images to map changes in forest cover, allowing researchers to take regional context (forest species, wildlife habitats, etc.) into account when performing this type of analysis. • Demonstrated that existing DL networks can be re-purposed to scrutinise satellite images of completely different geographies. • Devised a language agnostic framework that could assist future researchers in improving the reproducibility of their work. 1/2 ○ Redi-shop: Online web store - (Go-Lang, Redis, Kubernetes, Docker) • Designed and deployed an online web-store that implemented separate micro-services for orders, payments, stock, and users. • Implemented two-phase commit using SAGAs and the measured the performance difference between a Redis and Postgres back-end. • Conclusion - Even on a smaller scale, Redis was noticeably faster and handled load balancing far better than a more traditional Postgres back-end. Interests & Hobbies • Writer for the Christiaan Huygens student association magazine at TU Delft. • Organised a blood donation drive in coordination with the Rotary Club of B.S.Abdur Rahman University and raised funds for HelpAge India, an organisation that assists the elderly. • Passionate about music and take particular interest in western music. • Avid reader - enjoy fiction & non-fiction, attend book clubs and poetry readings. 2/2