Developed a centralized Kafka configuration management system for the Online Feature Store within the Meesho/BharatMLStack repository, focusing on standardizing environment variables and deployment processes. The work involved designing a structured KafkaConfig system and an environment-based generator to unify Kafka settings across all consumer components, ensuring consistent deployments in development, staging, and production environments. By migrating Kafka configuration from Orion to the online-feature-store, the project reduced configuration drift and operational overhead, improving the reliability of streaming feature delivery. The implementation leveraged Go and emphasized configuration management and environment variable handling, resulting in faster onboarding and streamlined maintenance for the engineering team.