Product

End-to-end use cases

Six real workflows that show how AI Enterprise Search fits into your daily operations.

Your company uses SharePoint for policies, Confluence for engineering docs, and file shares for templates. A new hire needs to find the vacation policy.

1
Connect data sources
Data Sources → Add Source. Connect SharePoint (Graph API credentials), Confluence (API token), and a file share (FTP/SMB). AES auto-discovers sites, spaces, and directories.
2
Crawl and index
Trigger a full crawl for each source. AES extracts text from PDFs, DOCX, PPTX, and HTML pages, chunks the content, generates embeddings, and stores everything in pgvector.
3
Search with natural language
The new hire types "what is our vacation policy?" in the search bar. AES retrieves relevant chunks, reranks them, and generates an AI answer citing the HR policy document from SharePoint.
4
Follow up in chat
They click "Continue in chat" and ask "how do I request time off?" The AI answers with a step-by-step guide, citing the SharePoint policy and the related Confluence page where HR documented the process.
Tip
Delta sync re-crawls only changed content. Set it to run daily so the knowledge base stays fresh without manual intervention.

2. Call center knowledge assist#

Your contact center uses Genesys Cloud. Agents need quick answers from product documentation and past call transcripts.

1
Connect Genesys Cloud
Add a Genesys Cloud data source with OAuth2 credentials. AES fetches conversations, recordings, and speech-to-text transcripts with speaker labels.
2
Embed the search widget
Add the AES widget to your agent desktop. Copy the embed code from Settings → Widget and paste it into your CRM portal HTML.
3
Agent searches mid-call
A customer asks about a product return. The agent types the question in the embedded widget. AES searches product docs and similar past conversations, returning a cited answer in seconds.
4
Programmatic integration
Use the Answer Snippets API to integrate AI answers into your Slack bot. Agents type /ask how do I process a return? and get a cited answer directly in Slack.

3. Saved searches for regulatory monitoring#

A compliance team needs to monitor when new content mentions specific regulatory topics across all company data sources.

1
Search and save
Search for "GDPR data retention requirements" and click Save. Name the saved search "GDPR retention monitoring" and set frequency to Daily.
2
Enable email alerts
Toggle alerts on for the saved search. The compliance officer will receive an email whenever new content matching the query is indexed.
3
Review new matches
When a new policy document is uploaded to SharePoint mentioning GDPR retention, the compliance officer gets an email. They click through to see the new content highlighted with a "NEW" badge.
4
Run Now for ad-hoc checks
Before a quarterly review, click Run Now on the saved search to get the current top matches with delta tracking showing what's new since the last check.

4. Engineering documentation hub#

Your engineering team's documentation lives across Confluence, a public docs site, and internal SharePoint. Finding the right API reference takes multiple searches across multiple tools.

1
Connect all sources
Add Confluence (engineering space), a Web Crawler (public docs site URL), and SharePoint (architecture docs library). Each connector handles authentication, pagination, and content extraction independently.
2
Configure chunking
AES uses code-aware chunking for technical content — preserving code blocks, API signatures, and markdown structure during text splitting.
3
Search across everything
A developer searches "how do I authenticate to the payments API?" AES finds the API reference from the docs site, the architecture decision from Confluence, and the implementation example from SharePoint.
4
White-label for your team
Brand the search portal with your engineering team's logo and colors via Settings → Branding. Make it feel like an internal tool, not a third-party product.

Your company has thousands of policy documents spread across SharePoint and file shares. Finding all details for a specific policy number means searching multiple folders and reading dozens of pages.

1
Index policy documents
Connect SharePoint document libraries and file shares containing policy documents. AES extracts text from PDF, DOCX, and scanned documents (with OCR).
2
Enable Document Expansion
Turn on Document Expansion in Settings → AI Assistant. When a query like "policy ABC-12345 coverage limits" hits multiple chunks from the same policy document, AES automatically pulls in the full document for complete context.
3
Search by policy number or topic
An underwriter searches "what are the exclusions for policy ABC-12345?" AES finds the policy, retrieves all relevant sections, and generates a comprehensive cited answer covering all exclusion clauses.
4
Explore entity connections
The Knowledge Graph shows linked entities — the policyholder, the underwriting agent, related claims, and referenced regulatory documents — giving the full picture without manual cross-referencing.

Your SharePoint and Confluence contain thousands of images — product photos, architecture diagrams, screenshots, and team photos. Finding the right image means scrolling through folders.

1
Enable Image Search
Turn on Image Search in Settings → AI Assistant. AES generates visual embeddings for every image it crawls using CLIP or Azure AI Vision.
2
Search with natural language
Search "network architecture diagram" or "team photo from the offsite" — AES matches images by visual meaning, not just filenames. Smart query preprocessing handles conversational phrasing like "show me all photos with people in a conference room".
3
AI-powered analysis
Every image gets an AI-generated caption, classification (photo, chart, diagram, screenshot), and semantic tags. Visible text is extracted via OCR. All metadata is searchable alongside the visual embeddings.
4
Filter and browse
Filter results by image type, data source, or tags. Preview images in the lightbox with full metadata, download originals, or open the source document.