As a Principal Software Development Engineer at Mediaocean, I lead large-scale projects that automate and integrate multi-source ads inventory data, and optimize converged media campaigns.
▪ Multi-Source Ads Inventory Automation and Integration
•
Orchestrated large-scale ETL processes for handling, analyzing, and curating inventory from Linear TV, CTV, and OTT data vendors (Nielsen, Vizio, Teletrax, Beeswax, Beachfront, etc.)
•
Seamlessly integrated data pipelines with GHA deployment and Kibana for real-time monitoring and tracking.
•
Integrated with event-driven alerts to proactively mitigate system downtime.
▪ Converged Media Optimization
•
Took ownership of inventory and viewership data management and processing, developing task-specific data structures for accurate campaign spot selection in optimization.
•
Significantly improved overall system response time by 50% through Snowflake SQL query optimization
•
Collaborated closely with frontend engineers to design and implement server-side services, RESTful APIs, incorporated with caching (Redis) and hybrid authentications in Agile style, sticking to CI/CD approach.
•
Proficient in utilizing Kubernetes for effective pod management and resource orchestration.
•
Successfully conducted proof of concept tasks to determine optimal implementation approach for reach & frequency data prediction.
▪ Specialized in mentoring and providing comprehensive technical guidance and walkthroughs, received positive feedback from 10+ junior backend engineers for clear and organized onboarding process, resulting in reduced ramp-up time and increased productivity.
▪ Received consistent accolades from direct manager and teammates for exceptional expertise in rapid bug debugging and turnaround.
Led the successful delivery of the Unified Measurement product within 3 months collaborated with the team. Leveraged highly abstracted class design for long term benefits and AWS services for effective management of large datasets.
•
Played a pivotal role in designing and implementing critical functionalities for the NBCUniversal’s One Platform, facilitating daily inventory data ingestion and enabling seamless delivery of specific features like scheduled measurement reports.
Implemented and delivered features for SCOPE TV Planner backend application and web server.
•
Led the migration effort to transition backend application code for four projects from Python 2 to Python 3, ensuring compatibility, improved performance, and adherence to modern development standards.- Served as the maintainer of the company's critical data warehousing, processing over 7k files 30GB data daily, and migrated Python 2 to 3 version changes with precision and accuracy.
•
Took ownership of an independent project as a new hire, driving its success and securing a $370K deal with a new data partner. Independently integrated a new data provider and client into our platform.
