{"id":16665,"date":"2025-12-03T16:53:56","date_gmt":"2025-12-03T15:53:56","guid":{"rendered":"https:\/\/www.panter.ch\/?p=16665"},"modified":"2025-12-03T16:53:56","modified_gmt":"2025-12-03T15:53:56","slug":"ai-powered-website-search","status":"publish","type":"post","link":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/","title":{"rendered":"How can AI help to better find relevant content on large websites?"},"content":{"rendered":"[vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;default&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221;][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h3>Table of contents<\/h3>\n[\/vc_column_text]<div class=\"nectar-fancy-ul link_style_underline nectar-link-underline-effect\" data-list-icon=\"fa fa-arrow-down\" data-animation=\"true\" data-animation-delay=\"0\" data-color=\"accent-color\" data-spacing=\"default\" data-alignment=\"left\"> \n<ul>\n<li><strong>The limits of classic navigation and the &#8220;trust paradox&#8221;<\/strong><\/li>\n<li><strong>What does the blueprint for an AI-supported website search look like?<\/strong><\/li>\n<li><strong>The strategic shift: from editor to curator<\/strong><\/li>\n<li><strong>Conclusion: Courage for new infrastructure<\/strong><\/li>\n<\/ul>\n <\/div>[\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;default&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221;][divider line_type=&#8221;Full Width Line&#8221; line_thickness=&#8221;1&#8243; divider_color=&#8221;default&#8221; animate=&#8221;yes&#8221;][\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; top_padding=&#8221;40px&#8221; top_padding_phone=&#8221;12vw&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;20px&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221; column_padding_type=&#8221;default&#8221; gradient_type=&#8221;default&#8221;][vc_column_text css_animation=&#8221;fadeInUp&#8221; css=&#8221;&#8221; text_direction=&#8221;default&#8221; max_width=&#8221;800&#8243;]\n<h3>In many companies and blogs, it is often the top 10 to top 20 pages that attract the majority of visitors. The rest only generate a fraction of the traffic. <!-- notionvc: b7e8cefc-ee2a-4adf-ae0c-df03b6c37730 --><!-- notionvc: 310ed2f7-764c-4f29-8a48-5cfb13b44cce --><\/h3>\n[\/vc_column_text][nectar_gradient_text heading_tag=&#8221;h3&#8243; color=&#8221;extra-color-gradient-1&#8243; gradient_direction=&#8221;horizontal&#8221; text=&#8221;An analysis by Ahrefs confirms this observation. According to this, around 33% of all search traffic on the company&#8217;s blog is accounted for by just five articles. &#8221; margin_top=&#8221;48&#8243; margin_bottom=&#8221;48&#8243;][image_with_animation image_url=&#8221;16668&#8243; image_size=&#8221;full&#8221; max_width=&#8221;100%&#8221; max_width_mobile=&#8221;default&#8221; animation_type=&#8221;entrance&#8221; animation=&#8221;None&#8221; animation_movement_type=&#8221;transform_y&#8221; hover_animation=&#8221;none&#8221; alignment=&#8221;&#8221; border_radius=&#8221;none&#8221; box_shadow=&#8221;none&#8221; image_loading=&#8221;default&#8221;][divider line_type=&#8221;No Line&#8221;][nectar_highlighted_text color_type=&#8221;regular&#8221; highlight_color=&#8221;#7000FF40&#8243; text_color=&#8221;#000000&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]Website managers of large organizations are familiar with this phenomenon as the <strong>Pareto principle<\/strong>: <em>20% of the pages generate 80% of the traffic<\/em>. But what happens to the remaining 80% &#8211; the countless subpages, PDFs and service descriptions? They form a <strong>&#8220;long tail&#8221;<\/strong> which, without a powerful AI search, becomes outdated over time, disappears in the information and navigation structure and ultimately becomes an invisible cost center.<br \/>\n<!-- notionvc: 2854c8be-af81-460f-a385-9e20e94f8491 -->[\/nectar_highlighted_text][nectar_highlighted_text color_type=&#8221;regular&#8221; highlight_color=&#8221;#7000FF40&#8243; text_color=&#8221;#000000&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]An example of the <a href=\"https:\/\/www.post.ch\/de\"><strong>Swiss Post<\/strong><\/a> shows the dimensions: <em>10,000 pages in four languages plus 5,500 images<\/em>. Depending on the average text volume per page, <strong>at least 6 full-time positions<\/strong> are required to maintain such a volume manually. To make this mass of information usable, we need to rethink the &#8220;website&#8221; concept &#8211; away from manual maintenance and towards an <strong>intelligent, AI-supported website search<\/strong>.<br \/>\n<!-- notionvc: 2854c8be-af81-460f-a385-9e20e94f8491 -->[\/nectar_highlighted_text][\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; top_padding=&#8221;40px&#8221; top_padding_phone=&#8221;12vw&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; id=&#8221;ch1&#8243; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;10px&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221; column_padding_type=&#8221;default&#8221; gradient_type=&#8221;default&#8221;][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h2>The limits of classic navigation and the &#8220;trust paradox&#8221;<\/h2>\n[\/vc_column_text][nectar_highlighted_text color_type=&#8221;regular&#8221; text_color=&#8221;#000000&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]Before an <strong>AI-powered website search<\/strong> can be implemented, it is important to understand why the status quo is failing.<br \/>\n<!-- notionvc: 2854c8be-af81-460f-a385-9e20e94f8491 -->[\/nectar_highlighted_text]<div class=\"nectar-fancy-ul\" data-list-icon=\"icon-salient-check\" data-animation=\"true\" data-animation-delay=\"0\" data-color=\"accent-color\" data-spacing=\"5px\" data-alignment=\"left\"> \n<ul>\n<li><strong>The navigation trap<\/strong> is a central problem of classic CMS and website structures. While CMS usually organize content hierarchically, human questions are associative, i.e. based on links between ideas. If the answer to a customer question is spread across several pages, a classic search bar will never put them together satisfactorily <span class=\"citation-294\">&#8211; a task that only a modern <\/span><b><span class=\"citation-294\">AI search function<\/span><\/b><span class=\"citation-294\"> can accomplish.<\/span><\/li>\n<li><strong>The &#8220;trust paradox&#8221;:<\/strong> A key learning from large-scale implementations (such as our <a href=\"https:\/\/www.panter.ch\/en\/portfolio\/post\/\">case study<\/a>) is user psychology. Users often trust a pure chatbot response less than a cleanly formatted website. A static page exudes <strong>&#8220;official validity&#8221;<\/strong>, while a chat window is often perceived as non-committal.<\/li>\n<\/ul>\n <\/div>[nectar_highlighted_text color_type=&#8221;regular&#8221; highlight_color=&#8221;#7000FF40&#8243; text_color=&#8221;#000000&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]The solution lies in an architecture that combines the <em>flexibility<\/em> of an LLM (Large Language Model) with the <em>authority<\/em> and UX of a classic website.<br \/>\n<!-- notionvc: 2854c8be-af81-460f-a385-9e20e94f8491 -->[\/nectar_highlighted_text][\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; top_padding=&#8221;40px&#8221; top_padding_phone=&#8221;12vw&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; id=&#8221;ch2&#8243; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;10px&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221; column_padding_type=&#8221;default&#8221; gradient_type=&#8221;default&#8221;][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h2>What does the blueprint for an AI-supported website search look like?<br \/>\n<!-- notionvc: 6c12655d-e77c-4ca4-847b-8e7987a79330 --><\/h2>\n[\/vc_column_text][nectar_highlighted_text color_type=&#8221;regular&#8221; text_color=&#8221;#000000&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]In order to realize such a solution, an architecture structured over four levels has proven itself for us:<br \/>\n<!-- notionvc: 456ec290-f57c-4074-b5b4-730b42297b7e --><\/p>\n<p><!-- notionvc: 2854c8be-af81-460f-a385-9e20e94f8491 -->[\/nectar_highlighted_text][divider line_type=&#8221;No Line&#8221;][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h3>Level 1: Semantic vectorization (from text to context)<br \/>\n<!-- notionvc: 6c12655d-e77c-4ca4-847b-8e7987a79330 --><\/h3>\n[\/vc_column_text][nectar_highlighted_text color_type=&#8221;regular&#8221; highlight_color=&#8221;#7000FF40&#8243; text_color=&#8221;#000000&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]The foundation of an <strong>AI-supported website search<\/strong> is the move away from keyword matching towards <em>semantic vectorization of<\/em> all content (HTML, PDFs, images). These are converted into vectors (embeddings) and stored in a vector database.<br \/>\n<!-- notionvc: 456ec290-f57c-4074-b5b4-730b42297b7e --><\/p>\n<p><!-- notionvc: 2854c8be-af81-460f-a385-9e20e94f8491 -->[\/nectar_highlighted_text][nectar_highlighted_text color_type=&#8221;regular&#8221; highlight_color=&#8221;#7000FF40&#8243; text_color=&#8221;#000000&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]<strong>The effect:<\/strong> an <em>on-site search with AI<\/em> can understand the semantic proximity, i.e. the meaning. Terms such as &#8220;parcel stamp&#8221; and &#8220;letter postage&#8221; are mathematically close to each other. If a user asks for &#8220;costs for shipping&#8221;, the system finds the relevant sections, even if the word &#8220;costs&#8221; does not appear there at all. The rigid CMS becomes a <strong>fluid knowledge network<\/strong>. <em>It is irrelevant where the information is located; only its context is important.<\/em><\/p>\n<p><!-- notionvc: 456ec290-f57c-4074-b5b4-730b42297b7e --><!-- notionvc: 2854c8be-af81-460f-a385-9e20e94f8491 -->[\/nectar_highlighted_text][\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; top_padding=&#8221;40px&#8221; top_padding_phone=&#8221;12vw&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; id=&#8221;ch2&#8243; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;10px&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221; column_padding_type=&#8221;default&#8221; gradient_type=&#8221;default&#8221;][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h3 data-start=\"355\" data-end=\"406\">Stage 2: The modular multi-LLM architecture of the AI search function<\/h3>\n[\/vc_column_text][nectar_highlighted_text color_type=&#8221;regular&#8221; highlight_color=&#8221;#7000FF40&#8243; text_color=&#8221;#000000&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]To solve the <strong>&#8220;Trust Paradox&#8221;<\/strong>, we rely on a <em>modular multi-LLM architecture<\/em> instead of a simple OpenAI wrapper. This approach offers several strategic advantages:<br \/>\n<!-- notionvc: 456ec290-f57c-4074-b5b4-730b42297b7e --><\/p>\n<p><!-- notionvc: 2854c8be-af81-460f-a385-9e20e94f8491 -->[\/nectar_highlighted_text]<div class=\"nectar-fancy-ul\" data-list-icon=\"icon-salient-check\" data-animation=\"true\" data-animation-delay=\"0\" data-color=\"accent-color\" data-spacing=\"5px\" data-alignment=\"left\"> \n<ul>\n<li><strong>Flexibility and independence:<\/strong> Reduced dependence on a single provider. It is possible to switch between models or combine them according to requirements.<\/li>\n<li><strong>Cost control and scalability:<\/strong> Costs can be saved by using local models for specific tasks and limiting API calls to specialized requirements.<\/li>\n<li><strong>Transparency and reproducibility:<\/strong> Critical or sensitive processes can be carried out using transparent models, while external LLMs are used for other tasks.<\/li>\n<li><strong>Optimization for specific use cases:<\/strong> The functionalities of LLMs can be improved by providing your own data and specialized prompts.<\/li>\n<\/ul>\n <\/div>[\/vc_column][\/vc_row][vc_row type=&#8221;full_width_background&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; bg_color=&#8221;#000000&#8243; scene_position=&#8221;center&#8221; top_padding=&#8221;80px&#8221; constrain_group_1=&#8221;yes&#8221; bottom_padding=&#8221;80px&#8221; bottom_margin=&#8221;40px&#8221; text_color=&#8221;light&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; class=&#8221;row dark&#8221; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;default&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221;][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h3 data-start=\"355\" data-end=\"406\">How exactly does this AI-supported website search work technically?<\/h3>\n[\/vc_column_text][nectar_highlighted_text color_type=&#8221;regular&#8221; text_color=&#8221;#FFFFFF&#8221; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]It is based on a multi-stage pipeline that strategically orchestrates several LLM calls:<br \/>\n<!-- notionvc: 456ec290-f57c-4074-b5b4-730b42297b7e --><\/p>\n<p><!-- notionvc: 2854c8be-af81-460f-a385-9e20e94f8491 -->[\/nectar_highlighted_text][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h4 data-start=\"355\" data-end=\"406\">Intent recognition and query expansion<\/h4>\n<p>The user input is first analyzed by an LLM to <strong>extract<\/strong> the <strong>intent (keyword\/topic)<\/strong>. Five semantically different search queries are then generated from the user prompt. This increases the hit rate and covers different formulations of the same question.[\/vc_column_text][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h4 data-start=\"355\" data-end=\"406\">Multi-Source Vector Search<\/h4>\n<p>The generated queries search the vector database for <strong>relevant content from various sources<\/strong> (websites, PDFs, images). The best results are filtered and provided as context for the next stage.[\/vc_column_text][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h4 data-start=\"355\" data-end=\"406\">Component selection through LLM<\/h4>\n<p>An LLM receives the filtered search results and selects the appropriate UI components from the design system to answer a question optimally. The system takes three important aspects into account when making this selection: <strong>Content complementarity, page layout and positioning, and content diversity.<\/strong>[\/vc_column_text][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h4 data-start=\"355\" data-end=\"406\">Component mapping and prompt assignment<\/h4>\n<p>Each selected component is <strong>mapped<\/strong> with a <strong>specific prompt<\/strong> that describes the requirements of this component.[\/vc_column_text][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h4 data-start=\"355\" data-end=\"406\">Parallel property generation<\/h4>\n<p>Now comes the decisive step. Each component runs through an individual LLM call. The following are transferred:[\/vc_column_text]<div class=\"nectar-fancy-ul\" data-list-icon=\"icon-salient-check\" data-animation=\"true\" data-animation-delay=\"0\" data-color=\"accent-color\" data-spacing=\"5px\" data-alignment=\"left\"> \n<ul>\n<li>The React component definition (available properties)<\/li>\n<li>The component-specific prompt<\/li>\n<li>The relevant content context<\/li>\n<\/ul>\n <\/div>[vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<p data-start=\"355\" data-end=\"406\">The LLM generates the specific <strong>properties and children<\/strong> for the React component. The advantages are<\/p>\n[\/vc_column_text]<div class=\"nectar-fancy-ul\" data-list-icon=\"icon-salient-check\" data-animation=\"true\" data-animation-delay=\"0\" data-color=\"accent-color\" data-spacing=\"5px\" data-alignment=\"left\"> \n<ul>\n<li>Granular control over each component<\/li>\n<li>Optimized, component-specific prompts<\/li>\n<li>Better error handling and debugging<\/li>\n<li>More efficient token usage<br \/>\n<!-- notionvc: b8d7defb-ead1-4d63-92ca-3995020334e3 --><\/li>\n<\/ul>\n <\/div>[vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h4 data-start=\"355\" data-end=\"406\">JSON assembly and client-side rendering<\/h4>\n<p>The validated component properties are sent <strong>to the client<\/strong> as <strong>structured JSON<\/strong>.[\/vc_column_text][\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; top_padding=&#8221;40px&#8221; top_padding_phone=&#8221;12vw&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; id=&#8221;ch3&#8243; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;10px&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221; column_padding_type=&#8221;default&#8221; gradient_type=&#8221;default&#8221;][vc_column_text css_animation=&#8221;fadeInUp&#8221; css=&#8221;&#8221; text_direction=&#8221;default&#8221; max_width=&#8221;700&#8243;]\n<h3>Stage 3: Generative hybrid UX for intuitive AI search<\/h3>\n<p><!-- notionvc: d8dacc4a-a524-4930-8edc-9a4ef7ab3761 -->[\/vc_column_text][nectar_highlighted_text color_type=&#8221;regular&#8221; highlight_color=&#8221;#7000FF40&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]The result for users is not a chat dialog, but a <em>dynamically generated landing page<\/em>. For example, if users ask for &#8220;insurance cover abroad&#8221;, the AI search generates a <strong>coherent, complete page in the corporate design<\/strong> on the fly that answers this exact question &#8211; including links and filter options. Users often don&#8217;t even notice that this page didn&#8217;t exist seconds ago.[\/nectar_highlighted_text][\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; top_padding=&#8221;40px&#8221; top_padding_phone=&#8221;12vw&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;10px&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221; column_padding_type=&#8221;default&#8221; gradient_type=&#8221;default&#8221;][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h3>Stage 4: Automated quality assurance (evals)<br \/>\n<!-- notionvc: d77fd744-8512-43a5-ba33-69153a4508c3 --><\/h3>\n[\/vc_column_text][nectar_highlighted_text color_type=&#8221;regular&#8221; highlight_color=&#8221;#7000FF40&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]In the corporate context, <strong>&#8220;AI hallucinations&#8221;<\/strong> are unacceptable. As manual approvals are impossible for dynamic content, we recommend <em>&#8220;unit tests&#8221; (evals) for content<\/em>. These work as follows:[\/nectar_highlighted_text]<div class=\"nectar-fancy-ul\" data-list-icon=\"icon-salient-check\" data-animation=\"true\" data-animation-delay=\"0\" data-color=\"accent-color\" data-spacing=\"5px\" data-alignment=\"left\"> \n<ul>\n<li>A checker (often a stronger model) compares the generated answer with the <strong>&#8220;ground truth&#8221;<\/strong> (the vectorized original documents).<\/li>\n<li>If the response deviates in fact or violates the tonality guidelines, for example, it is blocked or corrected before it reaches the user. This guarantees brand safety without manual effort.<\/li>\n<\/ul>\n <\/div>[\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; top_padding=&#8221;40px&#8221; top_padding_phone=&#8221;12vw&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; id=&#8221;ch4&#8243; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;10px&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221; column_padding_type=&#8221;default&#8221; gradient_type=&#8221;default&#8221;][vc_column_text css_animation=&#8221;fadeInUp&#8221; css=&#8221;&#8221; text_direction=&#8221;default&#8221; max_width=&#8221;700&#8243;]\n<h2>The strategic shift: from editor to curator<\/h2>\n<p><!-- notionvc: d8dacc4a-a524-4930-8edc-9a4ef7ab3761 -->[\/vc_column_text][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]The implementation of this blueprint changes an organization in the long term:[\/vc_column_text]<div class=\"nectar-fancy-ul\" data-list-icon=\"icon-salient-check\" data-animation=\"true\" data-animation-delay=\"0\" data-color=\"accent-color\" data-spacing=\"5px\" data-alignment=\"left\"> \n<ul>\n<li><strong>Single source of truth:<\/strong> The content team no longer maintains finished pages. They maintain tidbits of information and data. The AI takes care of the composition.<\/li>\n<li><strong>Scalability:<\/strong> Whether you add 100 or 10,000 documents does not affect performance or findability. The system grows organically.<\/li>\n<li><strong>Future-proof:<\/strong> Because all data is available in vectorized form, the front end works completely and independently.<\/li>\n<\/ul>\n <\/div>[\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;default&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221;][divider line_type=&#8221;Full Width Line&#8221; line_thickness=&#8221;1&#8243; divider_color=&#8221;default&#8221; animate=&#8221;yes&#8221;][\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; id=&#8221;ch5&#8243; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221; gradient_type=&#8221;default&#8221; shape_type=&#8221;&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;default&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221;][vc_column_text css_animation=&#8221;fadeInUp&#8221; css=&#8221;&#8221; text_direction=&#8221;default&#8221;]\n<h2 data-start=\"483\" data-end=\"553\">Conclusion: Courage for new infrastructure<\/h2>\n[\/vc_column_text][nectar_highlighted_text color_type=&#8221;regular&#8221; highlight_color=&#8221;#7000FF40&#8243; style=&#8221;full_text&#8221; highlight_expansion=&#8221;default&#8221; text_direction=&#8221;default&#8221;]An efficient <strong><span class=\"notion-enable-hover\" data-token-index=\"1\">AI-supported website search<\/span><\/strong> has long since become an IT architecture problem &#8211; not primarily an SEO problem. But beware: while AI systems remain dependent on traditional search engines, their integration requires complex technical requirements. Those who still manage floods of information manually will quickly be left behind &#8211; but those who implement an <strong>AI search function<\/strong> without creating solid foundations will also fail. <!-- notionvc: a2e848c3-664f-4d71-900b-c2d7e019b502 -->[\/nectar_highlighted_text][\/vc_column][\/vc_row][vc_row type=&#8221;in_container&#8221; full_screen_row_position=&#8221;middle&#8221; column_margin=&#8221;default&#8221; column_direction=&#8221;default&#8221; column_direction_tablet=&#8221;default&#8221; column_direction_phone=&#8221;default&#8221; scene_position=&#8221;center&#8221; text_color=&#8221;dark&#8221; text_align=&#8221;left&#8221; row_border_radius=&#8221;none&#8221; row_border_radius_applies=&#8221;bg&#8221; overflow=&#8221;visible&#8221; overlay_strength=&#8221;0.3&#8243; gradient_direction=&#8221;left_to_right&#8221; shape_divider_position=&#8221;bottom&#8221; bg_image_animation=&#8221;none&#8221;][vc_column column_padding=&#8221;no-extra-padding&#8221; column_padding_tablet=&#8221;inherit&#8221; column_padding_phone=&#8221;inherit&#8221; column_padding_position=&#8221;all&#8221; column_element_direction_desktop=&#8221;default&#8221; column_element_spacing=&#8221;default&#8221; desktop_text_alignment=&#8221;default&#8221; tablet_text_alignment=&#8221;default&#8221; phone_text_alignment=&#8221;default&#8221; background_color_opacity=&#8221;1&#8243; background_hover_color_opacity=&#8221;1&#8243; column_backdrop_filter=&#8221;none&#8221; column_shadow=&#8221;none&#8221; column_border_radius=&#8221;none&#8221; column_link_target=&#8221;_self&#8221; column_position=&#8221;default&#8221; gradient_direction=&#8221;left_to_right&#8221; overlay_strength=&#8221;0.3&#8243; width=&#8221;1\/1&#8243; tablet_width_inherit=&#8221;default&#8221; animation_type=&#8221;default&#8221; bg_image_animation=&#8221;none&#8221; border_type=&#8221;simple&#8221; column_border_width=&#8221;none&#8221; column_border_style=&#8221;solid&#8221;][vc_column_text css=&#8221;&#8221; text_direction=&#8221;default&#8221;]Do you have your own AI project in mind or would you like to find out how a customized AI solution can help your company? Contact us now.<br \/>\n<!-- notionvc: 6562536d-bf1f-4a74-85a6-2aa2c36a87f6 -->[\/vc_column_text][nectar_btn size=&#8221;large&#8221; button_style=&#8221;regular&#8221; button_color_2=&#8221;Accent-Color&#8221; icon_family=&#8221;none&#8221; text=&#8221;Contact us now&#8221; url=&#8221;https:\/\/www.panter.ch\/en\/contact\/&#8221;][\/vc_column][\/vc_row]\n","protected":false},"excerpt":{"rendered":"<p>Discover our blueprint for an AI-supported website search. This finally makes content on large websites easier to find. <\/p>\n","protected":false},"author":3,"featured_media":15074,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"om_disable_all_campaigns":false,"inline_featured_image":false,"footnotes":""},"categories":[276,358,274,361],"tags":[],"class_list":{"0":"post-16665","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-ai-data-en","8":"category-ai-applications","9":"category-blog-en","10":"category-llm-chatbots-en"},"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v27.3 (Yoast SEO v27.3) - https:\/\/yoast.com\/product\/yoast-seo-premium-wordpress\/ -->\n<title>AI-supported website search and how it works explained<\/title>\n<meta name=\"description\" content=\"Discover our blueprint for an AI-supported website search. Finally make content on large websites easier to find.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"How can AI help to better find relevant content on large websites?\" \/>\n<meta property=\"og:description\" content=\"Discover our blueprint for an AI-supported website search. Finally make content on large websites easier to find.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/\" \/>\n<meta property=\"og:site_name\" content=\"Panter AG\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/panterch\" \/>\n<meta property=\"article:published_time\" content=\"2025-12-03T15:53:56+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.panter.ch\/wp-content\/uploads\/HBC04064-1-1024x683.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1024\" \/>\n\t<meta property=\"og:image:height\" content=\"683\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"Roberto\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@panterch\" \/>\n<meta name=\"twitter:site\" content=\"@panterch\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Roberto\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"8 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/\"},\"author\":{\"name\":\"Roberto\",\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/#\\\/schema\\\/person\\\/17955a8092d4a5671347f2f158635479\"},\"headline\":\"How can AI help to better find relevant content on large websites?\",\"datePublished\":\"2025-12-03T15:53:56+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/\"},\"wordCount\":4159,\"publisher\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.panter.ch\\\/wp-content\\\/uploads\\\/HBC04064-1.jpg\",\"articleSection\":[\"AI &amp; Data\",\"AI Applications\",\"Blog\",\"LLM &amp; Chatbots\"],\"inLanguage\":\"en-US\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/\",\"url\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/\",\"name\":\"AI-supported website search and how it works explained\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.panter.ch\\\/wp-content\\\/uploads\\\/HBC04064-1.jpg\",\"datePublished\":\"2025-12-03T15:53:56+00:00\",\"description\":\"Discover our blueprint for an AI-supported website search. Finally make content on large websites easier to find.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/#primaryimage\",\"url\":\"https:\\\/\\\/www.panter.ch\\\/wp-content\\\/uploads\\\/HBC04064-1.jpg\",\"contentUrl\":\"https:\\\/\\\/www.panter.ch\\\/wp-content\\\/uploads\\\/HBC04064-1.jpg\",\"width\":2299,\"height\":1533,\"caption\":\"Zwei M\u00e4nner sitzen nebeneinander an einem Tisch, schauen auf einen Laptop-Bildschirm und diskutieren \u00fcber eine KI-L\u00f6sung. Beide scheinen engagiert und konzentriert zu sein, wobei einer in Richtung des Bildschirms gestikuliert. Im Hintergrund sind freiliegende Ziegelw\u00e4nde und andere arbeitende Menschen zu sehen.\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/stories\\\/ai-powered-website-search\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/salient-architect-landing\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"How can AI help to better find relevant content on large websites?\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/#website\",\"url\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/\",\"name\":\"Panter AG\",\"description\":\"Smarter, better, human.\",\"publisher\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/#organization\",\"name\":\"Panter AG\",\"alternateName\":\"Panter\",\"url\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/www.panter.ch\\\/wp-content\\\/uploads\\\/Panter_Favicon_696x696.jpg\",\"contentUrl\":\"https:\\\/\\\/www.panter.ch\\\/wp-content\\\/uploads\\\/Panter_Favicon_696x696.jpg\",\"width\":696,\"height\":696,\"caption\":\"Panter AG\"},\"image\":{\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/www.facebook.com\\\/panterch\",\"https:\\\/\\\/x.com\\\/panterch\",\"https:\\\/\\\/ch.linkedin.com\\\/company\\\/panter-ch\",\"https:\\\/\\\/www.instagram.com\\\/panter.ch\\\/\",\"https:\\\/\\\/github.com\\\/panter\"],\"description\":\"Wir von Panter sind Macher und Techies mit interdisziplin\u00e4rem Know-how \u2013 von Softwareentwicklung \u00fcber Design bis zu AI-Expertise. Mit innovativen, nutzerzentrierten L\u00f6sungen, modernster Technologie und einem starken Netzwerk entwickeln wir Services, die Ihre Nutzer begeistern und Ihr Unternehmen voranbringen. Als Schweizer Software- und Beratungsunternehmen stehen wir f\u00fcr Qualit\u00e4t, Innovationskraft und partnerschaftliche Zusammenarbeit \u2013 mit Herz und Verstand.\",\"email\":\"hello@panter.ch\",\"telephone\":\"+41 44 500 29 04\",\"legalName\":\"Panter AG\",\"foundingDate\":\"2005-01-01\",\"vatID\":\"CHE-262.202.358\",\"numberOfEmployees\":{\"@type\":\"QuantitativeValue\",\"minValue\":\"11\",\"maxValue\":\"50\"}},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/www.panter.ch\\\/en\\\/#\\\/schema\\\/person\\\/17955a8092d4a5671347f2f158635479\",\"name\":\"Roberto\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/63049a62ffa6445b337ce3fa0e7ad5dbf289f7541d5f9e0f10d871cc3b4241a8?s=96&d=mm&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/63049a62ffa6445b337ce3fa0e7ad5dbf289f7541d5f9e0f10d871cc3b4241a8?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/63049a62ffa6445b337ce3fa0e7ad5dbf289f7541d5f9e0f10d871cc3b4241a8?s=96&d=mm&r=g\",\"caption\":\"Roberto\"}}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"AI-supported website search and how it works explained","description":"Discover our blueprint for an AI-supported website search. Finally make content on large websites easier to find.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/","og_locale":"en_US","og_type":"article","og_title":"How can AI help to better find relevant content on large websites?","og_description":"Discover our blueprint for an AI-supported website search. Finally make content on large websites easier to find.","og_url":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/","og_site_name":"Panter AG","article_publisher":"https:\/\/www.facebook.com\/panterch","article_published_time":"2025-12-03T15:53:56+00:00","og_image":[{"width":1024,"height":683,"url":"https:\/\/www.panter.ch\/wp-content\/uploads\/HBC04064-1-1024x683.jpg","type":"image\/jpeg"}],"author":"Roberto","twitter_card":"summary_large_image","twitter_creator":"@panterch","twitter_site":"@panterch","twitter_misc":{"Written by":"Roberto","Est. reading time":"8 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/#article","isPartOf":{"@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/"},"author":{"name":"Roberto","@id":"https:\/\/www.panter.ch\/en\/#\/schema\/person\/17955a8092d4a5671347f2f158635479"},"headline":"How can AI help to better find relevant content on large websites?","datePublished":"2025-12-03T15:53:56+00:00","mainEntityOfPage":{"@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/"},"wordCount":4159,"publisher":{"@id":"https:\/\/www.panter.ch\/en\/#organization"},"image":{"@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/#primaryimage"},"thumbnailUrl":"https:\/\/www.panter.ch\/wp-content\/uploads\/HBC04064-1.jpg","articleSection":["AI &amp; Data","AI Applications","Blog","LLM &amp; Chatbots"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/","url":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/","name":"AI-supported website search and how it works explained","isPartOf":{"@id":"https:\/\/www.panter.ch\/en\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/#primaryimage"},"image":{"@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/#primaryimage"},"thumbnailUrl":"https:\/\/www.panter.ch\/wp-content\/uploads\/HBC04064-1.jpg","datePublished":"2025-12-03T15:53:56+00:00","description":"Discover our blueprint for an AI-supported website search. Finally make content on large websites easier to find.","breadcrumb":{"@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/#primaryimage","url":"https:\/\/www.panter.ch\/wp-content\/uploads\/HBC04064-1.jpg","contentUrl":"https:\/\/www.panter.ch\/wp-content\/uploads\/HBC04064-1.jpg","width":2299,"height":1533,"caption":"Zwei M\u00e4nner sitzen nebeneinander an einem Tisch, schauen auf einen Laptop-Bildschirm und diskutieren \u00fcber eine KI-L\u00f6sung. Beide scheinen engagiert und konzentriert zu sein, wobei einer in Richtung des Bildschirms gestikuliert. Im Hintergrund sind freiliegende Ziegelw\u00e4nde und andere arbeitende Menschen zu sehen."},{"@type":"BreadcrumbList","@id":"https:\/\/www.panter.ch\/en\/stories\/ai-powered-website-search\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.panter.ch\/en\/salient-architect-landing\/"},{"@type":"ListItem","position":2,"name":"How can AI help to better find relevant content on large websites?"}]},{"@type":"WebSite","@id":"https:\/\/www.panter.ch\/en\/#website","url":"https:\/\/www.panter.ch\/en\/","name":"Panter AG","description":"Smarter, better, human.","publisher":{"@id":"https:\/\/www.panter.ch\/en\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.panter.ch\/en\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.panter.ch\/en\/#organization","name":"Panter AG","alternateName":"Panter","url":"https:\/\/www.panter.ch\/en\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.panter.ch\/en\/#\/schema\/logo\/image\/","url":"https:\/\/www.panter.ch\/wp-content\/uploads\/Panter_Favicon_696x696.jpg","contentUrl":"https:\/\/www.panter.ch\/wp-content\/uploads\/Panter_Favicon_696x696.jpg","width":696,"height":696,"caption":"Panter AG"},"image":{"@id":"https:\/\/www.panter.ch\/en\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/panterch","https:\/\/x.com\/panterch","https:\/\/ch.linkedin.com\/company\/panter-ch","https:\/\/www.instagram.com\/panter.ch\/","https:\/\/github.com\/panter"],"description":"Wir von Panter sind Macher und Techies mit interdisziplin\u00e4rem Know-how \u2013 von Softwareentwicklung \u00fcber Design bis zu AI-Expertise. Mit innovativen, nutzerzentrierten L\u00f6sungen, modernster Technologie und einem starken Netzwerk entwickeln wir Services, die Ihre Nutzer begeistern und Ihr Unternehmen voranbringen. Als Schweizer Software- und Beratungsunternehmen stehen wir f\u00fcr Qualit\u00e4t, Innovationskraft und partnerschaftliche Zusammenarbeit \u2013 mit Herz und Verstand.","email":"hello@panter.ch","telephone":"+41 44 500 29 04","legalName":"Panter AG","foundingDate":"2005-01-01","vatID":"CHE-262.202.358","numberOfEmployees":{"@type":"QuantitativeValue","minValue":"11","maxValue":"50"}},{"@type":"Person","@id":"https:\/\/www.panter.ch\/en\/#\/schema\/person\/17955a8092d4a5671347f2f158635479","name":"Roberto","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/63049a62ffa6445b337ce3fa0e7ad5dbf289f7541d5f9e0f10d871cc3b4241a8?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/63049a62ffa6445b337ce3fa0e7ad5dbf289f7541d5f9e0f10d871cc3b4241a8?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/63049a62ffa6445b337ce3fa0e7ad5dbf289f7541d5f9e0f10d871cc3b4241a8?s=96&d=mm&r=g","caption":"Roberto"}}]}},"_links":{"self":[{"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/posts\/16665","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/comments?post=16665"}],"version-history":[{"count":4,"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/posts\/16665\/revisions"}],"predecessor-version":[{"id":16699,"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/posts\/16665\/revisions\/16699"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/media\/15074"}],"wp:attachment":[{"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/media?parent=16665"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/categories?post=16665"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.panter.ch\/en\/wp-json\/wp\/v2\/tags?post=16665"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}