Turn your knowledge base into a polished search experience your team or customers can use right away. The Interface tab lets you control how your search looks, behaves, and who can access it — with a live preview so you can see every change as you make it.
To access it, open a knowledge base and click the Interface tab in the top navigation.
The left side of the screen shows all configuration sections. The right side shows a live preview that updates as you make changes.
Basics
Shape what your users see when they search.
Max document results per search
Set how many document results appear per query (default: 6). More results give users more to explore; fewer results keep the experience focused and faster to scan.
Show relevancy score for every document result
Let users gauge how closely each result matches their query — useful for power users who want to evaluate result quality at a glance.
Re-ranked documents
Surface the most relevant results first by adding an extra ranking step after retrieval. Especially helpful for broad queries where the first pass might not perfectly order the results.
Style and personalisation
Make the search experience feel like part of your brand.
Accent color
Choose the accent color used throughout the search interface (buttons, highlights, links) using the color picker.
Display image
Upload your logo or avatar to appear in the search interface. Suggested size is 56 x 56 pixels. Supported formats: PNG, JPEG.
Display name
Set the name that appears at the top of the search interface (e.g., your company name).
Show welcome image
Toggle this on to greet users with a welcome image when they first land on the search page.
Welcome image
Upload the image that appears above the welcome message.
Search bar placeholder text
Guide your users with custom placeholder text (e.g., “Search our knowledge base…” or “Ask a question…”).
Enable initial prompts
Help new users get started by showing suggested questions before they type — this reduces the blank-page problem and shows what the search can do.
AI overview
Shape the quality and accuracy of AI-generated answers your users receive.
AI overview generation prompt
Define instructions for how the AI generates answers. For example: “Answer strictly based on the provided context; do not introduce information outside it, and respond clearly, concisely, and accurately within these limits.”
LLM provider for AI overview
Select the AI provider used to generate answers (e.g., OpenAI).
Model for AI overview
Select the specific model used to generate answers (e.g., gpt-4o-mini).
Max token per LLM call
Set the maximum number of tokens the AI can use per response. The default is 1000.
Advanced search retrieval
Get more accurate answers by adding an LLM-powered retrieval step that refines which chunks are used before generating the answer.
Advanced search retrieval mode
Balance speed and accuracy based on what your users need:
| Mode | When to use |
|---|
| fast | High-volume or real-time search where speed is critical |
| accurate | Precision-focused search where getting the right answer matters most |
Advanced search retrieval model
Select the model used for the advanced retrieval step (e.g., gpt-4.1-mini).
Start with the default AI overview prompt and adjust based on user feedback. If answers are too verbose, add instructions to be concise. If answers miss important details, widen the prompt to allow more thorough responses.
Security
Control who can access your search — keep it open for public use or restrict it to the right people.
Protect with SSO Auth
Ensure only authenticated team members can access the search by requiring SSO (Single Sign-On) login.
Protect with Password
Add a simple password gate — useful for sharing with clients or partners who don’t have SSO access.
Building an internal knowledge base? SSO Auth keeps it secure without any extra steps for your team. For a public-facing search, leave both toggles off.
Deploying your search
Your search is ready to go live. Share it as a direct link or embed it into any website.
Open Search
At the top of the Interface tab, you’ll see a URL like https://app.vectorshift.ai/search/deployed/.... — this is the direct link to your deployed search. Click Open Search to open it in a new tab, or use the copy button to share the URL with your team.
Embed Search
Add search directly to your website or app. Click Embed Search to get an iframe code snippet you can drop into any platform — Wix, Squarespace, Framer, Webflow, WordPress, or custom HTML.
The dialog shows the HTML iframe code with a Copy button.
Save
Click the Save button in the top-right corner to save your current interface configuration. The button shows a green “Saved” indicator once your changes are saved.
Always click Save after making changes. The live preview updates in real time, but your changes are not persisted until you save.
Applying search filters
Help your users get more targeted results by narrowing their search to specific files in the knowledge base.
Click Apply Filter in the search preview (or in the deployed search) to open the Filter modal. This modal shows the knowledge base contents, and you can select or deselect individual files to include or exclude them from the search scope.
Use the Search bar and All Statuses dropdown within the filter modal to quickly find specific files. Selected files are highlighted with a checked checkbox.
A note at the bottom of the modal reads: “Filters will be applied for your next search prompt.”
Click Apply to save the filter, or Cancel to discard.