Vipparthi Keerthi

Completed a paid internship in Machine Learning. Project: Modeling Planetary Atmospheres using ML (focused on Jupiter).

Use real user scenarios, simulate actual queries or workflows to see how prompts perform under real conditions. Monitor edge cases, watch for unexpected model behavior, latency spikes, or token overflows. Log and compare outputs, track differences between local and production response to catch environment specific quirks. Version prompts, keep a change log of prompt iterations and link them to test results. Manual sanity checks, critical prompts before every release to catch regressions.

Spearheaded automation initiatives to enhance data integrity and reduce manual intervention across multiple domains. Designed and developed a webbased platform (VA System) supporting Finance, HR, CAD/GIS, and Supply Chain Modules. Collaborated with Autodesk to deliver AutoCAD-integrated urban planning solutions. Authored and maintained QA documentation, ensuring clarity and consistency in internal processes. Created and executed detailed test scenarios, managed test data, and reported defects using jira. Conducted retesting and regression testing Beta, Sandbox, Live, and UAT environments. Facilitated cross-team communication to resolve production issues and improve platform performance.

Contributed to the implementation of POT, a web-based ERP tool for construction project management. Prepared detailed test scenarios and cases across multiple modules, ensuring robust quality assurance. Collaborated with cross-functional teams for root cause analysis and corrective action planning. Reported and tracked defects using BugZilla and Wrike, presenting summaries to stakeholders. Conducted retesting and regression testing across Beta, Sandbox, and Live Environments. Managed test data documentation to support effective and repeatable testing purposes. Demonstrated transferable skills in QA, defect analysis, and testing relevant to AI and data science roles.

Contributed to the development of Patagonia EMR, supporting Public Health, Mental Health and Family Practice workflows. Collaborated with cross-functional teams to analyze systems, gather requirements, and document client specific solutions. Optimized manual test cases using domain expertise, improving testing success across diverse EMR functionalities. Led platform transition efforts through rigorous testing, documentation, and user support for smooth practitioner onboarding. Resolved user issues via troubleshooting and ticket escalation, ensuring timely and effective resolutions. Enhanced testing efficiency by adapting manual scripts for automated feature validation.