Work Experience

LLM driven Unit Testcase Generation

#

Pioneered the development and implementation of a cutting-edge, codebase-agnostic solution for automated unit test case generation using Large Language Models (LLMs). This innovation significantly enhanced test coverage and streamlined the testing process across diverse large-scale firmware and software codebases, leading to substantial improvements in testing efficiency and reduced manual effort.

➜Achieved significant, quantifiable improvements, including 100% functional coverage and an increase in conditional coverage from 50% to 83%, with a 92% code acceptance rate for generated tests, demonstrating high quality and reliability.

➜ Engineered sophisticated LLM integration to generate high-quality unit test cases, including function mocks and stubs (e.g., KUnit-compatible), by leveraging advanced, dependency resolution for precise and effective test generation across various codebase languages and frameworks.

➜ Developed robust context provision and error handling mechanisms for LLM-based test generation: This included creating an innovative solution to efficiently retrieve function dependencies from massive codebases (1M+ files) and implementing LLM-driven automation for iterative correction of build/runtime errors to significantly improve test reliability.

➜ Designed and optimized an end-to-end iterative test generation and execution workflow: This seamlessly integrated automated test case generation, execution, comprehensive coverage reporting, and targeted generation for uncovered conditions, ultimately collaborating to directly improve overall testing efficiency and code quality.

O-RAN compliant M-Plane Test Solution.

#

Pioneered the development and automation of M-Plane test solutions, ensuring rigorous O-RAN specification conformance for a key client, within their cutting-edge Open RAN and 5G initiatives. Key contributions included:

➜ Developed and maintained 47 O-RAN compliant M-Plane test cases, directly contributing to the conformance test suite specifications. This involved:

➜ Developing and evolving Test Automation software within the OpenTAP framework using C#, establishing a robust platform for M-Plane testing and empowering users with custom test plan creation capabilities.

➜ Designed and implemented test cases within the M-Plane toolkit utilizing Python, enhancing testing versatility and scope.

➜ Implemented REST API integration within the Test Automation software to enable remote execution of M-Plane tests, facilitating distributed M-Plane testing, thereby improving test suite scalability.

➜ Streamlined reporting processes by automating TIFG report generation, meticulously capturing comprehensive test details including configuration, status, steps followed, and ODU/ORU responses. This automation significantly reduced manual effort and enabled efficient, automated reporting of test plan results.

➜ Guaranteed strict adherence to critical O-RAN standards, including O-RAN.WG4, Open Front Haul Conformance Test specifications, and M-Plane specifications, thereby ensuring the integrity and compliance of all developed test solutions.

During my internship at Truminds Software Systems, I had the opportunity to work on a project for a client. My role in the project involved software development using Python, specifically focusing on the implementation of Named Entity Recognition for name masking, multiprocessing and multi-threading, and web and social media scraping using Selenium.

Working on this project allowed me to expand my knowledge in software development using Python, specifically in the field of Named Entity Recognition, multiprocessing and multi-threading, and web scraping. I learned how to work with large datasets and implement advanced techniques to extract and process data from different sources. Additionally, I gained experience in working with Selenium, a web scraping tool, which helped me to develop my web scraping skills.

The experience helped me to develop my technical and professional skills and made me more confident in my abilities to work on real-world projects.