- Development related to migration of current stack to LTA stack with proper unit testing and integration testing.- Analysis for validation of correctness of the migration with big query and internal shadow testing validation tools.- Contributing to new feature requests and bug fixes in java and cpp legacy component.

- Spikes and pugh matrix evaluation on various open source rule engines, workflow engines for re-engineering the internal lending system more efficiently.- Built custom workflow engine library based on DAG to be used to better represent lending workflow with in-built support for pause resume through persistence - Automation of rules generation from the excel sheet by using expression parsing techniques like lexing and parsing with ANTLR. Also explored many other alternatives for ANTLR during spikes evaluation.- Built a dynamic in-memory lookup engine for efficiently performing lookup of crucial score cutoffs during the underwrting process. - Worked on identifying the root cause for performance issue in the fuzzy match engine being used in one of the project in production and fixed the same with around 50% decrease in response time.- Worked on identifying the root cause of a production issue due to failure in message queue connection causing failure of thousands of requests on a daily basis. This was due to tcp level buffer overflow, fixed the same by implementing best practices for connection establishment.- Automated jenkins deployment pipelines to execute all the deployment steps from groovy script which enabled code reusability, ease of use and maintainability.- Identified storage related issues arised in the fuzzy match-based data collection system and did 5x storage optimsation with optimisation of rdms schema. - Identified performance related issues arised in the fuzzy match-based data collection system. Identified the root cause being uneven clustering of data leading to I/O operation for every read of data. To mitigate the issue, suggested mongo-based caching for retrieval of widely distributed data and reduce the response time of slow requests within SLA.- Mentored juniors and college interns throughout the various phases of SDLC.- Identified potential race condition scenarios under high load and proposed a high level solution for the same.

Worked on building a system that is intended to gather all the customer information from various lending channels present in the company and store in a centralized manner. This centralized approach of storing data enables to extract more knowledge on customer data from all channels, paving way for making better business decisions.This system is built with a distributed web architecture approach. The system communicates with various lending channels and gets required customer information.My contributions include, - Writing gRPC and REST APIs for communicating with microservices, invoking CRUD operations, invoking algorithms, etc. - Writing JUnit test cases for ensuring the correctness of the implementation. - Setting up Jenkins pipelines for continuous integration and continuous deployment of the microservices through ansible with load balancer. - Load testing to check performance and scalability of the services to make them production-ready.- Writing client libraries for integration of these services with resilience and fallback mechanisms. - Involved in peer code review to learn about functionalities implemented by others and to give my technical inputs to improve code quality.Tech stack used: Java, Spring Boot, MySQL, Docker, Ansible, Gradle, Scala, Gatling.

Worked on enhancing a desktop application that was intended to manage an existing skill learning system present within the company in a more efficient way. There were several modules in the desktop application which included features like viewing courses, enrolling, scheduling courses, approval, notification. This application was meant to be used by users with different roles and responsibilities like admin, manager, team lead, employee. All these roles also had to be taken into consideration for every feature of the system. My contributions include, - Explore the existing code base and refactor the existing code base to use a DSL to handle sensitive databases credentials and query execution in a more secure way. - Implement functionality for the Admin role to add a new user, delete the existing user, modify an existing user, etc. which were previously done by Admin through MySQL command line. Tech stack used: C# with .NET framework, Windows forms, MySQL, Stored Procedures.