It’s been an eventful week in the open source community — if you’re a deep learning practitioner or Google Cloud customer, that is. A day after Amazon released Neo-AI, a framework for optimizing AI models, Google’s made the Cloud Search Connector SDK — a software development kit intended to bridge the gap between Google’s Cloud Search technology and enterprise content and identity repositories — freely available on GitHub.
As Google explained in a blog post, the identity and content connectors built with the SDK enable Google Cloud users to search on-premises, cloud, and software-as-a-service applications “more efficiently,” with the help of the Google Cloud Search Indexing API. Machine learning powers the API’s instant query suggestions and surfaces results across different content systems in over 100 different languages. Granular access-level controls — including individual-level, group-level, and content-based hierarchies — allow admins to prevent users from seeing search results they shouldn’t.
The Connector SDK is designed to handle tasks like service communication, multithreaded API operations, traversal strategies, connector configuration, error handling, and more, and it’s been used by partners to develop third-party connectors for over 50 sources, Google says. That includes repositories like Amazon S3, Box, and Microsoft OneDrive; collaboration solutions like Atlassian Confluence, Jira, and Microsoft Sharepoint; and sales, support, and Enterprise Resource Planning (ERP) platforms like Salesforce, SAP, and ServiceNow.
Currently, there are more than 80 connector options available in the Cloud Search Connector directory.
“At Google, we strongly believe in community-driven development. Developers and partners in our ecosystem play a pivotal role in how we architect our products,” Praveen Krishnakumar, product manager at Google Cloud Platform, and Tanmay Vartak, Google software engineer, wrote in their blog post. “We welcome your feedback and contributions, and are excited to see how you’ll use the SDK to build connectors for Cloud Search and extend the SDK to support additional functionality.”
In addition to the code, Google has published accompanying developer guides on how to create content connectors and identity connectors, as well as a tutorial demonstrating how a simple deployment of Cloud Search might work.