Sparsh Verma

Data Administrator at Fuller's, Smith & Turner

Optimisation & Modelling: Designed and deployed Python-based ML microservices processing 20,000+ daily transactions with 99.5% uptime, directly improving operating performance across 127 units. Big Data Pipelines: Owned end-to-end pipelines handling 4 TB of monthly data, utilizing Spark to reduce processing time by 33.3% and enable same-day model deployment. Revenue Protection: Applied advanced anomaly detection on 7+ million records, identifying critical insights that prevented $10,000+ in potential losses quarterly. Business Intelligence: Developed real-time KPI dashboards that improved operational efficiency by 42% and enabled data-driven decision-making for executive leadership. Leadership & Collaboration: Partnered with 20+ cross-functional teams to deliver production-ready ML systems within 15-day cycles, ensuring alignment with organizational objectives.

Summer Intern (App Development) at MP Police

Worked on a security-critical production application supporting real-world public-sector deployment. Redesigned geofence and location intelligence algorithms, improving accuracy by 86%. Reduced false alerts by 60% through enhanced validation, error handling, and system tuning.

Android App Developer at Deep Story (India)

Contributed to end-to-end development of cloud-integrated, ML-driven applications deployed in production environments. Worked with Python and Java backend services and AWS-hosted databases to improve data reliability, latency, and scalability. Optimized reinforcement learning models, improving prediction accuracy by ~20% through architecture tuning and performance analysis.