Known answers
Use questions your team can verify. Check that the answer reflects the source, rather than merely sounding plausible.
Use case / Company knowledge
Make company knowledge easier to use.
Connect selected documents to a compatible AI client. Evaluate whether your team can find useful answers, inspect the supporting sources, and respect access boundaries.
Test the questions that matter.
Missing information
Include questions the documents cannot answer. Assess whether the workflow makes that limitation clear.
Restricted sources
Check retrieval with different roles. Verify that users cannot access sources outside their intended permissions.
Know what the workflow needs.
Srasta supports ingestion of text-based PDF, DOCX, RTF, supported text files, and Git sources. A compatible client uses the retrieval and model APIs to provide the employee-facing experience.
Document ingestion does not imply OCR for scanned files or automatic synchronization of permissions from workplace applications. Review source configuration and the complete client data flow.
Read the ingestion guide →Retrieval today. Memory is planned.
Retrieval brings relevant source material into a task. Srasta Membrane is a separate planned direction for persistent, permission-aware context across work.
Learn about planned Membrane →Start with one team’s questions.
We can discuss source formats, access requirements, client fit, and success criteria in a demo. Guided pilots are optional and separately scoped.