At EZOTechnologies, I built Dhanda’s Python-based serverless video-rendering pipeline for an AI-powered marketing video platform. By tuning Lambda memory, CPU allocation and concurrency, I cut render time by about 79%.
I build five to six marketing video templates a month for Dhanda, using reusable FFmpeg automation for subtitles, overlays, audio mixing and brand-token substitution. I also built an AI workflow canvas and verified image and video renders end to end.
I shipped a serverless HTML-to-image service that returns S3-hosted PNGs, and hardened access control for AI and workflow APIs. I’ve also worked on asynchronous AI video-generation workflows and cloud-service integrations.
Previously, at Hexaware Technologies, I built Python automation, Flask REST APIs and backend microservices for McAfee’s cloud-based tools. At CRISIL Ltd., I developed Python and Spark data-analysis pipelines and automated API-based data collection.
