● Automated the manual process of assessing Parkinson’s motor symptoms by developing a diagnostic tool using MediaPipe, YOLOv8, Django, and SciPy, reducing analysis time by 60% and enhancing detection accuracy.

● Implemented advanced signal processing algorithms (e.g., Butterworth filtering, peak detection) to quantify metrics like amplitude, RMS velocity, and cycle duration, improving tremor detection precision.

● Designed an interactive React frontend for dynamic waveform visualization, enabling real-time region selection for analysis and automatic extraction of clinical movement metrics.

* Developed the most-utilized MCP tool in SRE org (21.6% usage among 50+ tools) integrating ServiceNow with natural language

querying and incident impact metrics (business errors, TPV impact, ownership attribution), plus contributed to Slack agent for team operations including BigQuery execution and FCI impact report generation

* Engineered enterprise-scale automation platform reducing endpoint onboarding by 99.7% (3 days → 5 minutes) through

Airflow orchestration, processing 1000+ endpoints and attributing 14.34M average daily errors in Tier 1 services across PayPal's global payment ecosystem (PayPal, Venmo, Braintree, Xoom)

➢ Deployed a self-hosted n8n marketing instance running 2 million executions per month on a $200 infrastructure, achieving $10,000 monthly savings.

➢ Developed a new version of CMA Reports, reducing churn from 30% to 12% and driving sales with 400 active subscriptions, generating $160,000 in monthly recurring revenue and enhancing long-term profitability.

➢ Developed an email sequences feature generating 1000 leads monthly, significantly reducing lead acquisition costs by $200,000 and generating $100,000 in revenue, streamlining marketing efficiency and increasing ROI.

➢ Refactored the handling of enquiries from 16 external portals, processing around 2,000/month efficiently by parsing emails from different sources, enhancing operational efficiency by 40%.

➢ Implemented a segmented backup strategy in PostgreSQL, using pg_sample to streamline database sampling, reducing development branch backup sizes from 6GB to 500MB—a reduction of approximately 92%. Optimized

➢ data obfuscation by breaking it into two stages, reducing the time from 1.5 hours to 40 minutes—a 56% decrease, significantly enhancing developer productivity.

➢ Optimized Next.js page delivery by storing server-rendered pages in MongoDB and caching with Cloudflare Cache API, reducing response times from 160ms to 40ms—a 75% improvement in site speed.

➢ Developed a microservice for programming assignments using Spring Boot, Groovy, and Netflix OSS (Zuul, Ribbon), integrating the Sphere Engine API, resulting in a 30% faster submission processing.

➢ Implemented storage of uploaded course materials in Google Cloud Storage (GCS), improving scalability and reducing storage costs by 25%.

➢ Implemented lint-staged, Husky, Prettier, and ESLint in a React application to ensure consistent code formatting, which led to a 50% reduction in formatting-related review comments and significantly accelerated the development cycle.

➢ Set up GitHub Actions to automate deployment of a React app to Vercel, and Dockerized a Spring Boot application for deployment to Google Cloud Run, reducing deployment time by 40% and improving CI/CD efficiency.