Laxmi Soundarya Pothineni

Project: LLM-Olympus

• Analyzed large volumes of data related to associate performance (quality, AHT, error rates) using SQL, Excel, and Power BI, enabling data-driven decision-making.
• Developed and implemented annotation defect bifurcation strategies in SQL and Excel, classifying errors into controllable versus uncontrollable, leading to more targeted process improvements.
• Reduced controllable errors by 25% through structured root cause analysis, and implementing targeted training interventions based on performance insights.
• Provided data-driven recommendations and UI enhancements that improved annotation efficiency by 15%, optimizing AHT, while maintaining quality and enhancing overall tool usability.
• Developed and tracked key performance metrics (Quality Score, AHT, Productivity, Error Rate) to drive continuous process improvements and align annotation efforts with AI model needs.

Project: Just Walk Out Technology

• Analyzed the data and root cause defects using proprietary tools.
• Also used SQL to run queries, and Excel to data mine the results of queries.
• Working with the Project Management and Business Operations Team to understand their requirements to deliver adequate results.
• Preparing the daily accuracy report of the third-party stores using Excel.
• Given many Gemba ideas for process improvement.
• Part of creating a navigation/decision tree, which helps associates easily understand the SOP.
• Supported team as an SME and Auditor by taking weekly calls to discuss the defects and SOPs.
• Handled weekly Manager’s calibration calls.
• Part of UAT to update and improve the tools and processes by providing the inputs and feedback.