RAHUL MUTHYAPU

Product: Gemini i18n Text, Trust & safety, Gen AI (from JAN 2023)

In my current role as a Data Strategist, I specialize in analyzing policy violation patterns in LLM responses using SQL, Python. I generate adversarial queries, assess user feedback on Gemini's responses. I ensure compliance with Google's policies by analyzing and improving LLM's responses. Here's a breakdown of my ongoing responsibilities:

1. Policy violation Loss pattern analysis of LLM's response using SQL, Python

2. Adversarial Query generation using python, LLM's and internal tools

3. User feedback analysis of Gemini's response using SQL, Python, looker studio

4. Google dashboard for User feedback and social monitoring(Twitter, facebook) creation and maintaining it.

5. Localization of Gemini (i18n) LLM's response, focused on JP, KO, EN, HI, FR, ES and more languages

Product: Scam Ads, Trust & safety, Google Shopping.

I am actively involved in executing the company's strategy for enforcing shopping policies and addressing abuse within Google Shopping Products. This entails close collaboration with stakeholders to ensure the strict adherence of shopping policies and the prevention of fraud, spam, and abuse. Here's a breakdown of my responsibilities:

1. I investigate and resolve Merchant complaints related to the company's shopping policies, using data analysis and database queries to understand issues and find solutions.
2. I am actively engaged in projects aimed at reducing fraudulent activity on Google Shopping. This includes the development of algorithms and the identification of patterns of abuse, which requires programming and data analysis skills.
3. I provide continuous data analysis and propose solutions to drive decisions for improving productivity. This involves using my expertise in data analytics tools, statistical analysis, and data visualization on an ongoing basis.
4. I consistently develop metrics and dashboards using SQL to extract valuable insights from large-scale data, showcasing my proficiency in database management and data visualization software.
5. I collaborate with cross-functional teams to provide ongoing data analysis and reporting, which involves strong communication skills and the ability to interpret data effectively.
6. I have established rules that automatically identify non-compliant merchants using patterns and my programming skills that flag non-compliant behavior.

My current role demands a combination of technical and analytical skills to effectively enforce LLM's response, reduce abuse, and maintain a secure Gemini's environment to all users.

• Built out the data and reporting infrastructure from the ground up using Power BI and Azure to provide real-time insights into the Bugs, Testcases execution, feature executions, future work item for newly releasing Windows 11 Builds SV2, SV1, SV, and Hailey Windows 11 Build
• Used Azure ADO and Azure Boards for bug-tracking
• Developed root cause reports to address problems like Bug stats, feature priority with testing team conversions, successfully revealing insights that boosted conversions by 32%
• Built operational reporting in Power BI to find areas of improvement for Microsoft Clients resulting in decreasing 900 person hours
• Led a team of 2 full-time employees
• 5 Star rated by Microsoft for consecutive two quarters.

• Applied models and data to understand and predict material vendors for IoT devices on the Electrolux's market's list, and presented findings to stakeholders
• Built data visualizations using SQL and Power BI for Electrolux's vendor KPIs that reduced manual reporting work by 2 hours weekly, with a team of 4

• Familiar with NumPy, Pandas, Seaborn and Matplotlib
• Good experience in working with various Python Integrated Development Environments like Idle, and Jupyter Notebook
• Responsible in creating a database and its validations and authorization with MySQL
• Created database using MySQL, wrote several queries to extract data from database.