In February 2026, Shirshanka developed semantic search integration for the acryl-data/datahub-helm repository, enabling vector similarity search for entity discovery within DataHub deployments. Leveraging Go and YAML, Shirshanka introduced multi-provider embedding support for OpenAI and AWS Bedrock, centralized configuration under a global block, and implemented environment variable controls for flexible deployment. The work included refactoring provider configuration for maintainability, updating documentation, and enhancing credential management by securely wiring embedding credentials into the actions pod. These changes improved search relevance, onboarding, and security, demonstrating depth in API integration, Helm chart development, and Kubernetes-based cloud service orchestration within a production environment.