As a Software Engineer at ServiceNow, I play a key role in building intelligent solutions that enhance customer service, developer workflows, and instance health management through AI and automation. I work on the Impact product and several forward-looking AI initiatives, combining software engineering with applied machine learning and generative AI.

𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗔𝗜 – 𝗖𝗼𝗻𝘃𝗲𝗿𝘀𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗼𝗻 𝗳𝗼𝗿 𝗦𝗲𝗿𝘃𝗶𝗰𝗲𝗡𝗼𝘄 𝗖𝘂𝘀𝘁𝗼𝗺𝗲𝗿𝘀

𝗟𝗟𝗠-𝗕𝗮𝘀𝗲𝗱 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿 𝗧𝗼𝗼𝗹𝘀 & 𝗖𝘂𝘀𝘁𝗼𝗺 𝗦𝗼𝗹𝘂𝘁𝗶𝗼𝗻𝘀

Built an assistant that reviews code from GitHub pull requests using LLMs, helping developers catch issues and improve quality. Developed a web UI to visualize feedback and integrate seamlessly with dev workflows.

Created a plugin that allows developers to select and review code directly in VS Code using LLMs fine tuned for servicenow codebase.

Experimented with fine-tuning open-source LLMs using proprietary code, running and evaluating them locally via Ollama to ensure performance and domain adaptation.

During my SDE internship at Alexa’s Multitasking team, I innovatively designed and implemented a solution for session handling to enhance efficiency and reduce costs related to DB reads, writes, and EC2 host resources.

Utilizing Java, Spring Boot, Google Guice, and DynamoDB, I developed an in-memory session management system. This strategic move cut DynamoDB interactions significantly, resulting in a 10% annual cost reduction for database operations, contributing to competitive pricing for Alexa devices.

I leveraged AWS services, specifically EC2, to ensure scalability and reliability of our solution. GitHub was integral for version control, while JUnit and various mocking techniques were employed for rigorous testing, ensuring robustness in our implementation.

Collaborating closely with senior developers, I immersed myself in the software development life cycle, gaining hands-on experience and enhancing my skills in data structures and efficient coding practices. My role required a strong understanding of JSON for effective data interchange between services.

In terms of data mining, I applied clustering algorithms to categorize user sessions, facilitating pattern recognition and anomaly detection. This data-driven strategy bolstered continuous improvements in Alexa’s session handling, ensuring an optimal user experience.

In sum, my internship was a blend of software development, cloud computing, and data mining, leading to enhanced session management efficiency and significant cost savings for the team.

I served as an Instructional Student Assistant for a course on Data Mining, where my responsibilities encompassed developing comprehensive and challenging assignments aimed at fortifying students’ understanding of pivotal data mining concepts. My involvement in the course was extensive, ensuring that students had the resources and support needed to grasp complex topics and apply them in practical settings.

One of my primary tasks was creating assignments that delved into Exploratory Data Analysis (EDA). This involved guiding students through processes such as data cleaning, normalization, and visualization, ensuring they developed proficiency in identifying patterns, anomalies, and crucial insights from datasets.

Furthermore, I crafted assignments focused on various Machine Learning models, including Supervised Learning techniques like Decision Trees, Random Forests, and Support Vector Machines, as well as Unsupervised Learning methods such as K-Means Clustering and Hierarchical Clustering. I ensured that students not only understood the theoretical aspects of these models but also gained hands-on experience in implementing them.

In addition to traditional Machine Learning models, I incorporated Neural Networks into the curriculum, guiding students through the intricacies of designing and training these models. This included delving into various Activation Functions such as ReLU, Sigmoid, and Tanh, and elucidating their roles in the context of Deep Learning.

To further enhance students’ learning experience, I introduced assignments that explored different hardware acceleration techniques. This covered the utilization of Graphics Processing Units (GPUs) and Tensor Processing Units (TPUs) to expedite the training process of complex models, providing insights into the practical aspects of deploying Machine Learning and Deep Learning models in real-world scenarios.

As the Initiation Officer at Tau Beta Pi, the esteemed engineering honor society, I dedicated myself to promoting academic excellence and fostering a community of high-achieving engineers. My role involved scrutinizing academic records to identify and recruit top-tier students who demonstrated outstanding scholastic achievement and a strong grasp of engineering fundamentals.

Employing data analytics, I streamlined the candidate selection process, ensuring that only those who met our rigorous GPA and academic standards were considered. This approach not only upheld the society’s prestigious reputation but also maintained the integrity of our membership.

I was responsible for orchestrating the initiation ceremonies, a critical platform to instill the society’s values and emphasize the importance of academic integrity. My position required effective communication and coordination with faculty, alumni, and society members to ensure seamless event execution and uphold our long-standing traditions.

To enhance the professional development of our members, I initiated a series of technical workshops and seminars. These events leveraged the expertise of industry veterans and alumni, providing valuable insights into emerging trends, innovation, and best practices across various engineering disciplines.

Additionally, I maintained meticulous records of member achievements and society activities, applying data management principles to safeguard the confidentiality and accuracy of this information.

In sum, my tenure as Initiation Officer was marked by a commitment to excellence, fostering a nurturing environment that championed academic achievement, innovation, and a relentless pursuit of engineering excellence.

Worked as a consultant for Apple GBI Team. During this phase, I spearheaded the optimization of the order processing workflow, utilizing a robust Kafka and Apache Spark-based data pipeline. This high-performance solution resulted in a remarkable 40% reduction in processing time, ensuring timely and efficient order fulfillment.

To assist the Apple sales team in data management, I developed an end-to-end data administration tool. Employing React JS for the frontend, Java for backend operations, and SQL for database interactions, the tool led to a 30% decrease in data administration time, boosting the team's decision-making speed and overall responsiveness.

In a data-driven initiative, I designed a LSTM (Long Short-Term Memory) model to predict no-shows at Apple store appointments. With an impressive 90% accuracy, the model significantly enhanced resource allocation and subsequently improved the customer experience.

Git was my choice for source code management, paired with Jenkins to establish a continuous integration/continuous deployment (CI/CD) pipeline for the Kafka and Apache Spark-based system. This setup ensured rapid, secure, and efficient releases, maintaining a steady flow of high-quality code.

My role also involved engineering middleware components using Java, which served as the linchpin for seamless integrations between the React JS-based web application and the SQL database, enabling real-time data retrieval and manipulation.

Facilitating team collaboration and communication, I conducted daily scrum stand-up meetings, providing a platform for early feedback and transparent discussion.

Utilizing Spring Boot, AWS (EC2, S3), Scala, Postman, Docker, and SQL alongside the previously mentioned technologies, I ensured a comprehensive and high-performance solution, integrating data mining techniques such as anomaly detection and pattern recognition to further enhance system efficiency and accuracy.