Saranraj Gandhi

Detail-oriented, organized and meticulous Software Engineer skilled in application development, testing and optimization. Excels at coordinating ground-up planning, programming and implementation for core modules. Maintains strong object-oriented and software architecture fundamentals.

Machine Learning - Guvi Geek Networks, IITM Research Park

SEDGE - Predictive AI Analytics Platform:

• As a Python developer for SEDGE, an AI platform, I contributed to the development of comprehensive suite of AI tool. SEDGE enables users to perform various AI-related tasks, including datasets uploading, sampling, target selection, statistical modeling, visual analytics and predictive analytics (auto ML and custom ML). My role involved implementing functionalities for data preprocessing, applying statistical function and queries, feature engineering, and custom machine learning algorithms developments. I also developed 30 interactive charts for visual analytics. Additionally, I integrated all top Machine learning algorithms to provide users with predictive insights and useful recommendations.

DQMS - Data Quality Management System:

• For the NYK ship management client, I was part of the team that developed DQMS, an interactive and dynamic application for monitoring ship sensors data. The system receives sensors data from ships, storing it in an AWS S3 bucket. We implemented batch processing jobs triggered based on the ship's data receiving time. Once the data is processed, we perform logical analysis and calculations to create interactive dashboards, summaries, and analytical reports. These allow immediate monitoring and action-taking for the health of ships.

Support AI:

• Support AI is a support automation project designed to reduce human resources by 80 % by automating L3 support for clients. Users can load their JIRA data and train an LSTM model to predicts the functional team for each tickets. Additionally, the system predicts the type of fix, and root cause based on previous similar issues. We implemented a chatbot using LLM, Open AI, and LLAMA-2 to provide user assistance. If user encounter difficulties with application functionalities, they can chat with the bot, which replies based on the trained user support documents.

I-Nautix:

• I worked as a Python developer on an I-Nautix project, an application enabling ship pilots and audit managers worldwide to conduct vessel inspections and screenings. I was responsible for developing a data synchronization module, which involved transferring data from the client server to the I-Nautix database using ad hoc APIs. Additionally, I created functionality to send calculated and summarized data from I-Nautix to client's Azure directory using SAS URL.