CloudFerro has introduced Search Agents & Knowledgebases to Sherlock, making it easier to build AI assistants that answer questions based on an organization's own documents.

The new capability eliminates the need to build and maintain a Retrieval-Augmented Generation (RAG) pipeline from scratch. Instead of integrating multiple components for document processing, semantic search and retrieval, users can upload their documents and let Sherlock prepare them for AI-powered search.

Search Agents & Knowledgebases support PDF documents and presentations. During ingestion, Sherlock automatically performs OCR, splits documents into searchable chunks, generates embeddings and indexes the content in a Knowledgebase.

Once the Knowledgebase is ready, it can be connected directly to a Search Agent. The solution can be tested through Sherlock's user interface or Playground and integrated into production applications using the API.

The new capability is designed for organizations building AI assistants for their own knowledge and documentation.

Typical use cases include:

• internal knowledge assistants,
• product documentation chatbots,
• technical documentation search,
• Q&A over reports and policies,
• research assistants based on publications.

For organizations operating in regulated environments, data sovereignty remains a key differentiator. The entire solution – including models, Knowledgebases and data storage – runs on CloudFerro's European infrastructure, with data encrypted at rest.

Search Agents & Knowledgebases simplify one of the most complex parts of building enterprise AI assistants. Instead of assembling and maintaining a RAG pipeline, teams can focus on their applications while Sherlock handles document processing and semantic search behind the scenes

— said Jan Szypulski, Product Manager at CloudFerro

Search Agents & Knowledgebases are currently available in Public Preview. User feedback collected during this phase will help shape the General Availability (GA) release. Compatibility with the OpenAI Responses API is already on the product roadmap.

Go to CloudFerro Sherlock.