Data Analyst
Security Compliance: In accordance with production security standards, all API access configurations have been completely decoupled from the source tree. The Gemini token layers run exclusively via secure machine environment variables (os.getenv). Local Sandbox Execution: The backend repository architecture is designed to initialize a local state cache directory (/stored_vault_files) on application startup to preserve and mount files dynamically. Evaluation Framework: To review the full dual-track vision pipeline and intent search mechanics, please ensure a local environment variable key is configured on the testing host terminal before running the Uvicorn server wrapper. Omni Identity treats a student’s credentials not as isolated, dusty PDF files, but as a living, interconnected ecosystem. By using Gemini to dynamically read both raw text and flat screenshots, map them onto a visual dependency graph, and let users chat with their data using natural human queries, we have built a bulletproof, intuitive verification hub that turns data management into a seamless user experience.
To conduct a data-driven investigation into historical and modern IPL delivery records in order to determine whether traditional strategic biases, such as the perceived match-winning advantage of the coin toss, hold statistical validity, and to identify the specific match phases and elite player performances that dictate overall team success in the T20 format. Process Data-Driven workflow executed in Jupyter Notebook using Pandas, Matplotlib, and Seaborn: Exploratory Data Analysis: Profiled over 289,000 ball-by-ball IPL records using Pandas to verify data types, check dataset structure, and map missing values. Hypothesis Testing: Analyzed the correlation between winning the coin toss and match outcomes to expose "The Toss Myth," comparing tactical decisions vs. a true 50-50 win reality. Phase Impact Aggregation: Isolated the Powerplay, Middle, and Death overs to compute true run-rate (RPO) splits and run gaps. Data Visualization: Engineered polished, custom dark-themed charts using Seaborn and Matplotlib for deployment. Results Toss Advantage: 51.55% (Statistical Illusion) Chasing (Fielding First) Success: 54.70% Win Rate Defending (Batting First) Success: 45.43% Win Rate Middle Overs Margin: +7.78 Run Gap (Highest of all phases) Winning Death Over Tempo: 11.27 Runs Per Over (RPO) 2022-2026 Batting Leader: Shubman Gill (2,827 runs) 2022-2026 Bowling Leader: Yuzvendra Chahal (90 wickets) Reflection What I would do differently: Integrate External Factors: I would merge the ball-by-ball dataset with external parameters such as stadium dimensions, pitch types (dry, green, dusty), weather conditions, and dew factors to see how much they heavily skew the "toss decision" advantage.
This is V’s work on Wooble.
Build a profile that shows what you can do — and share it anywhere.