Talk To Listen

What I did

• Processed and analyzed a 3M+ token dataset using Python and SQL, applying exploratory data analysis and model evaluation to improve response accuracy to 85%.
• Developed end-to-end data and analytics pipelines for ingestion, inference tracking, monitoring, and optimization, reducing latency by 75% and infrastructure costs by 82%.