Store Knowledge Base

AI & Compliance

Store Knowledge Base

This page lets staff add store-specific reference information and ask AI questions about it.

Store Knowledge Base illustration

How to find this page

  1. 1

    Sign in to EliteHMS

  2. 2

    Choose the correct location from the top bar

  3. 3

    Select My Knowledge Base from the left menu

My Knowledge Base option in the left menu
My Knowledge Base in the left menu

My Documents

The My Documents tab lists uploaded or pasted documents. Use Search documents to find one by its name, medicine name, or description, or use All Categories to filter by category.

The summary cards show document count, documents processing, searchable vectors, and storage used. The storage bar shows how many document slots and how much space remain.

My Documents tab with summary cards and storage bar
Summary cards and the storage bar

Tip:If the document and storage counts are zero, no store documents have been added yet.

Add a document

  1. 1

    Select Add Document

  2. 2

    Upload a file, or choose Paste Text to add written content directly

  3. 3

    Provide a document name and choose a category

  4. 4

    Add optional details

    Such as medicine name, tags, source, and description.

  5. 5

    Select Upload & Vectorize to submit the document

Add Document form with name, category, and file upload
The Add Document form

Choose one of these categories:

  • Medicine Info: Information about medicines your pharmacy stocks.
  • Our Services: Clinic services and packages.
  • Our Protocols: Your clinic’s procedures.
  • Guidelines: Clinical guidelines your team follows.
  • Store Information: Pharmacy details, FAQs, and policies.
  • Pricing & Packages: Service fees and packages.
  • General: Other reference material.

The limits are:

  • Files: PDF, DOCX, TXT, MD, and JSON, up to 5 MB each.
  • Pasted text: up to 5 MB.
  • Each store: up to 25 documents and 15 MB total storage.

Tip:For a medicine document, entering the medicine name helps AI find relevant information. Keep the content clear and current; for example, use headings and include the effective date on policies or fee schedules.

Document status

  • Processing: The document is being prepared for search.
  • Ready: The document is searchable.
  • Failed: Read the error shown beside it, correct the source file or content, and upload it again.
Document list with Processing, Ready, and Failed statuses
Document statuses in the list

Note:A ready document cannot be edited from this list. Delete it and upload the corrected version if you need to replace it. Deleting also removes its searchable vectors.

Ask AI

  1. 1

    Choose the Ask AI tab

  2. 2

    Choose a question type

    See the two types below.

  3. 3

    Use a suggested question or type your own, and submit

  4. 4

    Review the answer and its source labels

  5. 5

    Open View retrieved chunks

    This shows the pieces of documents used for the answer.

  • General Query: Ask about your store documents, services, protocols, and other reference information.
  • Medicine Lookup: Enter a medicine name, then ask a related question, such as where the store’s document discusses its information.
Ask AI tab with General Query, Medicine Lookup, and suggested questions
The Ask AI tab

The page describes a three-stage search:

  1. It searches your store’s documents.
  2. Then it searches a global clinical database.
  3. It may also use the AI model’s general training knowledge.

The answer’s source label indicates which sources contributed.

Warning:If no store document is used, the answer may not reflect your store’s own policy or stock details.

Good information to add

Useful documents include current opening hours, service fees, store policies, medicine reference notes, and internal procedures.

Warning:Avoid uploading information that your team should not make available to the AI query workflow. Keep patient-identifying details out of general reference documents.

Short definitions

  • Vector: A searchable representation of a small piece of a document that helps AI find relevant information.
  • Processing: The app is preparing the content for search.
  • Vectorize: Convert document content into searchable pieces for AI retrieval.
  • Source: The document or database that contributed information to an answer.
  • Global clinical database: Clinical reference information shared beyond your store’s own documents.