{"id":224422,"date":"2026-03-03T06:54:11","date_gmt":"2026-03-03T06:54:11","guid":{"rendered":"https:\/\/www.9senses.ai\/?page_id=224422"},"modified":"2026-03-16T10:59:12","modified_gmt":"2026-03-16T10:59:12","slug":"chatbot-audit-framework","status":"publish","type":"page","link":"https:\/\/www.9senses.ai\/de\/chatbot-audit-framework\/","title":{"rendered":"Chatbot Audit Framework"},"content":{"rendered":"<div class=\"et_pb_section_0 et_pb_section et_section_regular et_block_section preset--group--divi-section--divi-box-shadow--default preset--group--divi-section--divi-sizing--default\">\n<div class=\"et_pb_row_0 et_pb_row et_block_row preset--group--divi-row--divi-box-shadow--default preset--group--divi-row--divi-sizing--default\">\n<div class=\"et_pb_column_0 et_pb_column et_pb_column_4_4 et-last-child et_block_column et_pb_css_mix_blend_mode_passthrough preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_text_0 et_pb_text et_pb_bg_layout_light et_pb_module et_block_module preset--module--divi-text--9f6bba78-428e-4c83-b816-460811c2c529\"><div class=\"et_pb_text_inner\"><h4>GenAI Audit Framework<\/h4>\n<\/div><\/div>\n<\/div>\n<\/div>\n\n<div class=\"et_pb_row_1 et_pb_row et_pb_row_3-4_1-4 et_block_row et_block_row_3-4_1-4 preset--group--divi-row--divi-box-shadow--default preset--group--divi-row--divi-sizing--default\">\n<div class=\"et_pb_column_1 et_pb_column et_pb_column_3_4 et_block_column et_pb_css_mix_blend_mode_passthrough preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_text_1 et_pb_text et_pb_bg_layout_light et_pb_module et_block_module preset--module--divi-text--0241502a-1974-4cc3-b28c-915e7d98c0df\"><div class=\"et_pb_text_inner\"><h1>Generative AI<br \/>Audit Framework<\/h1>\n<\/div><\/div>\n\n<div class=\"et_pb_text_2 et_pb_text et_pb_bg_layout_light et_pb_module et_block_module preset--module--divi-text--0431a145-bb8b-4440-b3c5-0a16c179fb90\"><div class=\"et_pb_text_inner\"><p>The 9Senses GenAI Audit Framework is a structured methodology to assess how well a generative AI system performs against real business use cases \u2014 and how safely, responsibly, and reliably it behaves from a user, governance, and risk perspective. The framework combines use-case adherence, business outcomes, and standardized quality dimensions to produce an actionable improvement roadmap.<\/p>\n<p>The framework supports two audit levels:<strong> Level 1<\/strong> (behavioral \/ black-box evaluation based on observable behavior) and <strong>Level 2<\/strong> (open-book diagnosis covering architecture, retrieval\/RAG pipelines, agentic orchestration, governance, compliance, ethics, and value modeling). It applies across the full spectrum of GenAI deployments \u2014 customer-facing copilots, internal assistants, document processing systems, and multi-step autonomous agents.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_row_2 et_pb_row et_pb_row_nested et_flex_row preset--group--divi-row--divi-box-shadow--default preset--group--divi-row--divi-sizing--default\">\n<div class=\"et_pb_column_2 et_pb_column et_flex_column et_pb_css_mix_blend_mode_passthrough et_flex_column_12_24 et_flex_column_12_24_tablet et_flex_column_24_24_phone et_flex_column_12_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_text_3 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><h2>Level 1 Audit<\/h2>\n<p>An independent, external review of your chatbot's observable performance (black-box). No internal access to your technical environment required. The Level 1 audit is delivered within 5 business days.<\/p>\n<p><a href=\"\/de\/chatbot-audit\/#audit\">\u2192Book your Level 1 Audit here<\/a><\/p>\n<\/div><\/div>\n<\/div>\n\n<div class=\"et_pb_column_3 et_pb_column et-last-child et_flex_column et_pb_css_mix_blend_mode_passthrough et_flex_column_12_24 et_flex_column_12_24_tablet et_flex_column_24_24_phone et_flex_column_12_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_text_4 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><h2>Level 2 Audit<\/h2>\n<p>Eine ma\u00dfgeschneiderte vertiefte Analyse, z.B. der technischen Architektur, Retrieval-Systematik, Governance, Compliance und Business Value. Unbedingt empfohlen, wenn Level 1 gravierende Probleme zutage f\u00f6rdert.<\/p>\n<p><a href=\"\/de\/chatbot-audit\/#contact\">\u2192 Contact us to learn more\u00a0<\/a><\/p>\n<\/div><\/div>\n<\/div>\n<\/div>\n\n<div class=\"et_pb_heading_0 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h2 class=\"et_pb_module_header\">Why a GenAI Audit?<\/h2><\/div><\/div>\n\n<div class=\"et_pb_text_5 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><p>Generative AI systems are increasingly embedded in high-stakes workflows \u2014 customer service, sales, legal, HR, and operations \u2014 where failures translate directly into reputational, regulatory, or financial risk. A structured<span>\u00a0<\/span><strong>GenAI audit<\/strong><span>\u00a0<\/span>surfaces weaknesses before they cause harm, and validates that the system actually delivers its intended use cases and measurable business value.<\/p>\n<ul>\n<li><strong>Use-case adherence:<\/strong><span>\u00a0<\/span>Does the system reliably complete the tasks it was built for?<\/li>\n<li><strong>Business outcomes:<\/strong><span>\u00a0<\/span>Does it improve containment, conversion, resolution time, cost efficiency, or satisfaction?<\/li>\n<li><strong>Risk exposure:<\/strong><span>\u00a0<\/span>Does it hallucinate, leak sensitive data, or fail under adversarial prompts?<\/li>\n<li><strong>Governance readiness:<\/strong><span>\u00a0<\/span>Are transparency signals, human-in-the-loop controls, escalation paths, and monitoring adequate?<\/li>\n<li><strong>Regulatory compliance:<\/strong><span>\u00a0<\/span>Does it meet applicable obligations (EU AI Act, GDPR, sector-specific rules)?<\/li>\n<\/ul>\n<\/div><\/div>\n\n<div class=\"et_pb_heading_1 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h2 class=\"et_pb_module_header\">Level 1: Behavioral (black-box) Audit<\/h2><\/div><\/div>\n\n<div class=\"et_pb_text_6 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><p><strong>Level 1<\/strong> is a standardized external evaluation based on publicly observable behavior at the time of testing. It benchmarks performance, identifies user-facing weaknesses, and delivers a prioritized set of recommendations \u2014 without requiring access to the underlying system.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_text_7 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><h3>Core Dimensions &amp; Weights<\/h3>\n<div class=\"table-wrap\">\n<table>\n<thead>\n<tr>\n<th>Dimension<\/th>\n<th>What is evaluated<\/th>\n<th>Weight<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><strong>Output Quality<\/strong><\/td>\n<td>Relevance to the user's goal, factual correctness, completeness, formatting, hallucination robustness, citation behavior<\/td>\n<td><strong>50%<\/strong><\/td>\n<\/tr>\n<tr>\n<td><strong>Responsiveness<\/strong><\/td>\n<td>Latency for simple and complex requests, streaming behavior, perceived speed under load<\/td>\n<td><strong>20%<\/strong><\/td>\n<\/tr>\n<tr>\n<td><strong>User Interface &amp; Experience<\/strong><\/td>\n<td>Clarity, usability, layout, readability, affordance design, accessibility signals<\/td>\n<td><strong>15%<\/strong><\/td>\n<\/tr>\n<tr>\n<td><strong>Interaction &amp; Dialog Quality<\/strong><\/td>\n<td>Conversation flow, expectation management, tone calibration, multi-turn coherence, escalation behavior<\/td>\n<td><strong>15%<\/strong><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p>Additional dimensions are reviewed as<span>\u00a0<\/span><strong>outside-in indicators<\/strong><span>\u00a0<\/span>(not part of the Level 1 composite score):<span>\u00a0<\/span><strong>Business Value<\/strong>,<span>\u00a0<\/span><strong>Compliance<\/strong>, and<span>\u00a0<\/span><strong>Ethics<\/strong>. These inform the qualitative narrative and flag areas requiring deeper Level 2 investigation.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_heading_2 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h3 class=\"et_pb_module_header\">Scoring Methodology<\/h3><\/div><\/div>\n\n<div class=\"et_pb_text_8 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><p>Each dimension is scored on a standardized 1\u20135 scale and aggregated using the weights above into an overall performance score.<\/p>\n<\/div><\/div>\n<\/div>\n\n<div class=\"et_pb_column_4 et_pb_column et_pb_column_1_4 et-last-child et_block_column et_pb_css_mix_blend_mode_passthrough preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_image_0 et_pb_image et_pb_module et_block_module preset--group--divi-image--divi-box-shadow--default preset--group--divi-image--divi-sizing--default\"><span class=\"et_pb_image_wrap\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/www.9senses.ai\/wp-content\/uploads\/2024\/02\/ai-generator-8.png\" width=\"553\" height=\"408\" srcset=\"https:\/\/www.9senses.ai\/wp-content\/uploads\/2024\/02\/ai-generator-8.png 553w, https:\/\/www.9senses.ai\/wp-content\/uploads\/2024\/02\/ai-generator-8-480x354.png 480w\" sizes=\"(min-width: 0px) and (max-width: 480px) 480px, (min-width: 481px) 553px, 100vw\" class=\"wp-image-81\" title=\"ai-generator-8\" \/><\/span><\/div>\n\n<div class=\"et_pb_image_1 et_pb_image et_pb_module et_flex_module preset--group--divi-image--divi-box-shadow--default preset--group--divi-image--divi-sizing--default\"><a href=\"\/wp-content\/uploads\/2026\/03\/2026_02_28_Audit-Example_D.pdf\" target=\"_blank\"><span class=\"et_pb_image_wrap\"><img loading=\"lazy\" decoding=\"async\" src=\"\/wp-content\/uploads\/2026\/02\/2026_02_28_Audit-Example_D_Page_1-scaled.jpg\" title=\"2026_02_17_Audit Example ENG_Page_1\" width=\"724\" height=\"1024\" srcset=\"\" sizes=\"(min-width: 0px) and (max-width: 480px) 480px, (min-width: 481px) 724px, 100vw\" class=\"wp-image-224330\" data-srcset=\"\" \/><\/span><\/a><\/div>\n\n<div class=\"et_pb_text_9 et_pb_text et_pb_bg_layout_light et_clickable et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><p><a href=\"\/wp-content\/uploads\/2026\/02\/2026_02_28_Audit-Example_D.pdf\" target=\"_blank\" rel=\"noopener\">Please click to see a full Level 1 sample audit with methodology<\/a><\/p>\n<\/div><\/div>\n<\/div>\n<\/div>\n\n<div class=\"et_pb_row_3 et_pb_row et_flex_row preset--group--divi-row--divi-box-shadow--default preset--group--divi-row--divi-sizing--default\">\n<div class=\"et_pb_column_5 et_pb_column et_flex_column et_pb_css_mix_blend_mode_passthrough et_flex_column_1_5 et_flex_column_12_24_tablet et_flex_column_24_24_phone et_flex_column_12_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_heading_3 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h3 class=\"et_pb_module_header\">1<\/h3><\/div><\/div>\n\n<div class=\"et_pb_text_10 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><div class=\"score-item\">\n<p>Critical deficiency with high user impact or material risk<\/p>\n<\/div>\n<div class=\"score-item\"><\/div>\n<\/div><\/div>\n<\/div>\n\n<div class=\"et_pb_column_6 et_pb_column et_flex_column et_pb_css_mix_blend_mode_passthrough et_flex_column_1_5 et_flex_column_12_24_tablet et_flex_column_24_24_phone et_flex_column_12_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_heading_4 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h3 class=\"et_pb_module_header\">2<\/h3><\/div><\/div>\n\n<div class=\"et_pb_text_11 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><div class=\"score-item\">\n<p>Significant weaknesses requiring remediation<\/p>\n<\/div>\n<div class=\"score-item\"><\/div>\n<\/div><\/div>\n<\/div>\n\n<div class=\"et_pb_column_7 et_pb_column et_flex_column et_pb_css_mix_blend_mode_passthrough et_flex_column_1_5 et_flex_column_12_24_tablet et_flex_column_24_24_phone et_flex_column_12_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_heading_5 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h3 class=\"et_pb_module_header\">3<\/h3><\/div><\/div>\n\n<div class=\"et_pb_text_12 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><div class=\"score-item\">\n<p>Acceptable performance with identifiable limitations<\/p>\n<\/div>\n<div class=\"score-item\"><\/div>\n<\/div><\/div>\n<\/div>\n\n<div class=\"et_pb_column_8 et_pb_column et_flex_column et_pb_css_mix_blend_mode_passthrough et_flex_column_1_5 et_flex_column_12_24_tablet et_flex_column_24_24_phone et_flex_column_12_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_heading_6 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h3 class=\"et_pb_module_header\">4<\/h3><\/div><\/div>\n\n<div class=\"et_pb_text_13 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><div class=\"score-item\">\n<p>Strong performance with minor gaps<\/p>\n<\/div>\n<\/div><\/div>\n<\/div>\n\n<div class=\"et_pb_column_9 et_pb_column et-last-child et_flex_column et_pb_css_mix_blend_mode_passthrough et_flex_column_1_5 et_flex_column_12_24_tablet et_flex_column_24_24_phone et_flex_column_12_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_heading_7 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h3 class=\"et_pb_module_header\">2<\/h3><\/div><\/div>\n\n<div class=\"et_pb_text_14 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><div class=\"score-item\">\n<p>Best-practice level:<br \/>benchmark-worthy <br \/>performance<\/p>\n<\/div>\n<\/div><\/div>\n<\/div>\n<\/div>\n\n<div class=\"et_pb_row_4 et_pb_row et_pb_row_3-4_1-4 et_block_row et_block_row_3-4_1-4 preset--group--divi-row--divi-box-shadow--default preset--group--divi-row--divi-sizing--default\">\n<div class=\"et_pb_column_10 et_pb_column et_pb_column_3_4 et_block_column et_pb_css_mix_blend_mode_passthrough preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_text_15 et_pb_text et_pb_bg_layout_light et_pb_module et_block_module preset--module--divi-text--0431a145-bb8b-4440-b3c5-0a16c179fb90\"><div class=\"et_pb_text_inner\"><p>The 9Senses GenAI Audit Framework is a structured methodology to assess how well a generative AI system performs against real business use cases \u2014 and how safely, responsibly, and reliably it behaves from a user, governance, and risk perspective. The framework combines use-case adherence, business outcomes, and standardized quality dimensions to produce an actionable improvement roadmap.<\/p>\n<p>The framework supports two audit levels:<strong> Level 1<\/strong> (behavioral \/ black-box evaluation based on observable behavior) and <strong>Level 2<\/strong> (open-book diagnosis covering architecture, retrieval\/RAG pipelines, agentic orchestration, governance, compliance, ethics, and value modeling). It applies across the full spectrum of GenAI deployments \u2014 customer-facing copilots, internal assistants, document processing systems, and multi-step autonomous agents.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_heading_8 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h2 class=\"et_pb_module_header\">Level 2: Open-Book Audit (Root-Cause Diagnosis)<\/h2><\/div><\/div>\n\n<div class=\"et_pb_text_16 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><p><strong>Level 2<\/strong> is an open-book audit that explains why the GenAI system behaves as it does. It reviews the technical and organizational system behind the deployment and produces a targeted optimization plan grounded in root causes \u2014 not just symptoms.<\/p>\n<\/div><\/div>\n<\/div>\n\n<div class=\"et_pb_column_11 et_pb_column et_pb_column_1_4 et-last-child et_block_column et_pb_column_empty et_pb_css_mix_blend_mode_passthrough preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\"><\/div>\n<\/div>\n\n<div class=\"et_pb_row_5 et_pb_row et_flex_row preset--group--divi-row--divi-box-shadow--default preset--group--divi-row--divi-sizing--default\">\n<div class=\"et_pb_column_12 et_pb_column et-last-child et_flex_column et_pb_css_mix_blend_mode_passthrough et_flex_column_24_24 et_flex_column_24_24_tablet et_flex_column_24_24_phone et_flex_column_24_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_text_17 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><h3>Open-Book Scope &amp; Outcomes<\/h3>\n<div class=\"table-wrap\">\n<table>\n<thead>\n<tr>\n<th>Area<\/th>\n<th>What we review (examples)<\/th>\n<th>Typical outcomes<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><strong>Architecture &amp; Orchestration<\/strong><\/td>\n<td>System boundaries, tool\/API usage, routing logic, fallback strategy, agentic loops, trust zones, dependency risks<\/td>\n<td>Architecture risk map, refactoring recommendations, safer routing &amp; fallback design<\/td>\n<\/tr>\n<tr>\n<td><strong>Retrieval \/ RAG Quality<\/strong><\/td>\n<td>Chunking strategy, retrieval relevance, grounding behavior, citation logic, context window management, stale content risk<\/td>\n<td>Retrieval tuning plan, grounding improvements, measurable answer-quality lift<\/td>\n<\/tr>\n<tr>\n<td><strong>Prompting &amp; Guardrails<\/strong><\/td>\n<td>System prompts, policy hierarchy, refusal strategy, tool-use permissions, instruction injection robustness, output filters<\/td>\n<td>Hardened prompts, safer policies, reduced jailbreak &amp; prompt injection risk<\/td>\n<\/tr>\n<tr>\n<td><strong>Security &amp; Data Protection<\/strong><\/td>\n<td>Access controls, secrets handling, data minimization, leakage scenarios, PII exposure, logging sensitivity<\/td>\n<td>Risk remediation plan, control improvements, safer data flows<\/td>\n<\/tr>\n<tr>\n<td><strong>Governance &amp; Operations<\/strong><\/td>\n<td>Ownership model, human-in-the-loop design, escalation paths, monitoring, incident response, evaluation cadence, change management<\/td>\n<td>Operating model, monitoring design, continuous improvement loop<\/td>\n<\/tr>\n<tr>\n<td><strong>Compliance &amp; Ethics<\/strong><\/td>\n<td>Transparency &amp; disclosure obligations, data processing mapping, fairness considerations, EU AI Act risk classification, regulated use cases<\/td>\n<td>Compliance readiness checklist, documentation &amp; policy improvements<\/td>\n<\/tr>\n<tr>\n<td><strong>Business Value Modeling<\/strong><\/td>\n<td>KPI definitions, baseline vs. target, measurement instrumentation, token\/run-cost controls, ROI sensitivity analysis<\/td>\n<td>Value case, KPI dashboard spec, cost controls, prioritized roadmap<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<\/div><\/div>\n\n<div class=\"et_pb_heading_9 et_pb_heading et_pb_module et_flex_module preset--group--divi-heading--divi-box-shadow--default preset--group--divi-heading--divi-sizing--default\"><div class=\"et_pb_heading_container\"><h2 class=\"et_pb_module_header\">Methodology Summary<\/h2><\/div><\/div>\n\n<div class=\"et_pb_text_18 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><ul class=\"summary-list\">\n<li><strong>Level 1:<\/strong><span>\u00a0<\/span>structured behavioral testing (black-box), reproducible baseline scoring, prioritized recommendations \u2014 no system access required<\/li>\n<li><strong>Level 2:<\/strong><span>\u00a0<\/span>open-book root-cause diagnosis covering architecture, retrieval\/RAG, agentic orchestration, governance, compliance, and value modeling \u2014 produces a targeted optimization roadmap<\/li>\n<li><strong>Always:<\/strong><span>\u00a0<\/span>use-case adherence and business outcomes serve as the primary anchor for defining what \"good\" means in each context<\/li>\n<li><strong>Applies to:<\/strong><span>\u00a0<\/span>customer-facing AI assistants, internal copilots, document intelligence systems, multi-agent pipelines, and hybrid human+AI workflows<\/li>\n<\/ul>\n<\/div><\/div>\n<\/div>\n<\/div>\n\n<div class=\"et_pb_row_6 et_pb_row et_flex_row preset--group--divi-row--divi-box-shadow--default preset--group--divi-row--divi-sizing--default\" id=\"faq\">\n<div class=\"et_pb_column_13 et_pb_column et-last-child et_flex_column et_pb_css_mix_blend_mode_passthrough et_flex_column_24_24 et_flex_column_24_24_tablet et_flex_column_24_24_phone et_flex_column_24_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\">\n<div class=\"et_pb_text_19 et_pb_text et_pb_bg_layout_light et_pb_module et_flex_module preset--group--divi-text--divi-box-shadow--default preset--group--divi-text--divi-sizing--default preset--module--divi-text--default\"><div class=\"et_pb_text_inner\"><h2>FAQs - H\u00e4ufige Fragen und Antworten<\/h2>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_0 et_pb_accordion et_pb_module et_flex_module preset--group--divi-accordion--divi-box-shadow--default preset--group--divi-accordion--divi-sizing--default\">\n<div class=\"et_pb_accordion_item_0 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_open et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Was ist ein Chatbot Audit?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Ein Chatbot Audit ist eine strukturierte Bewertung des Verhaltens eines Chatbots. Es kann als Black-Box-Audit (Level 1) durchgef\u00fchrt werden, bei dem ausschlie\u00dflich das beobachtbare Verhalten analysiert wird, oder als Open-Box-Audit (Level 2), bei dem zus\u00e4tzlich Value-Generierung, Strukturen und Verhalten auf Basis detaillierter technischer Einblicke bewertet werden.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_1 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Warum ist ein Chatbot Audit sinnvoll?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Chatbots sind h\u00e4ufig der erste Kontakt f\u00fcr Interessenten und Kunden und beeinflussen direkt Kundenerfahrung, Markenwahrnehmung und operative Effizienz. Ein Chatbot-Audit identifiziert Schw\u00e4chen, bevor daraus Reputations- oder Gesch\u00e4ftsrisiken entstehen. Es definiert den Startpunkt und zeigt die Richtung f\u00fcr Optimierungspotenziale auf.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_2 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Wer sollte ein Chatbot Audit in Betracht ziehen?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p data-start=\"1502\" data-end=\"1812\">Organisationen, die KI-gest\u00fctzte Chatbots f\u00fcr Kundenservice, Vertrieb, Onboarding, Support oder interne Mitarbeiteranwendungen einsetzen, sollten ein Chatbot-Audit ins Auge fassen.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_3 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Welchen Umfang hat ein Level 1 Chatbot Audit?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Das Level-1-Chatbot-Audit umfasst strukturierte Use-Case-Tests, Halluzinations-Stresstests, Analysen von Weiterleitungs- und Eskalationsverhalten, Beobachtung der Dialogf\u00fchrung sowie optional eine Pr\u00fcfung der Mehrsprachigkeit.<\/p>\n<p>Ein Chatbot wird anhand folgender Dimensionen bewertet: Antwortqualit\u00e4t (50 % Gewichtung), Geschwindigkeit (20 %), Benutzeroberfl\u00e4che (15 %) und Dialogqualit\u00e4t (15 %). Daraus ergibt sich eine Gesamtnote.<\/p>\n<p><a href=\"\/wp-content\/uploads\/2026\/02\/2026_02_28_Audit-Example_D.pdf\" target=\"_blank\" rel=\"noopener\">Hier finden Sie ein vollst\u00e4ndiges Auditbeispiel.<\/a><\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_4 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Welche Methodik wird im Level 1 Chatbot Audit angewendet?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Das Audit baut auf dem 9senses KI Audit Framework auf. Es umfasst praxisnahe Funktionstests, Edge-Case-Szenarien (z. B. mehrdeutige oder ung\u00fcltige Eingaben), Halluzinations-Stresstests und Konsistenzpr\u00fcfungen.<\/p>\n<p>Jede Dimension wird auf einer standardisierten Skala von 1\u20135 bewertet und zu einer Gesamtnote aggregiert, um Vergleichbarkeit und Objektivit\u00e4t sicherzustellen.<\/p>\n<p><a href=\"\/wp-content\/uploads\/2026\/02\/2026_02_28_Audit-Example_D.pdf\" target=\"_blank\" rel=\"noopener\">Weitere methodologische Erl\u00e4uterungen finden Sie im Auditbeispiel.<\/a><\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_5 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Wie werden Halluzinationen im Level 1 Audit erkannt?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Halluzinationen \u2013 d.h. die Erzeugung unzutreffender oder erfundener Inhalte \u2013 stellen ein erhebliches Reputations- und Compliance-Risiko dar. Unser Audit beinhaltet gezielte Halluzinations-Stresstests.<\/p>\n<p>Dabei werden bewusst ung\u00fcltige Referenzen, Tippfehler und mehrdeutige Eingaben eingebracht, um zu pr\u00fcfen, ob der Chatbot Informationen erfindet oder r\u00fcckversichernde Nachfragen stellt. Bewertet werden Entit\u00e4tsvalidierung, Grounding-Verhalten und Eskalationslogik.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_6 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Pr\u00fcft das Chatbot Audit auch auf Compliance (EU AI Act, DSGVO)?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Level 1 beinhaltet eine erste Pr\u00fcfung offer, erkennbarer Indikatoren (KI-Kennzeichnung, Transparenzelemente, Datenschutzhinweise) sowie die Barrierefreiheit der Seite, auf welcher der Bot eingebunden ist.<\/p>\n<p>Eine vollst\u00e4ndige regulatorische und Governance Analyse \u2013 einschlie\u00dflich Dokumentations- und Architekturpr\u00fcfung \u2013 kann als Bestandteil eines Level 2 Chatbot Audits durchgef\u00fchrt werden.<\/p>\n<p class=\"not-prose mt-0! mb-0! flex-auto truncate\">\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_7 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Ist das Level 1 Chatbot Audit eine technische Pr\u00fcfung?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Nein. Das Level-1-Chatbot-Audit ist eine verhaltensbasierte Black-Box-Bewertung. Es analysiert beobachtbares Systemverhalten aus Nutzer- und Governance-Perspektive, ohne interne Architektur, Trainingsdaten, Retrieval-Systeme oder die Sicherheitsinfrastruktur zu pr\u00fcfen.<\/p>\n<p>Im Level 1 Audit berichten wir auch \u00fcber technische  Aspekte die wir aufgrund des beobachteten Verhaltens erkennen k\u00f6nnen. Technische Detailanalysen k\u00f6nnen Teil des Level-2-Audits sein.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_8 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Welche Informationen werden f\u00fcr einen Chatbot Audit ben\u00f6tigt?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p data-start=\"4233\" data-end=\"4511\">F\u00fcr Level 1 ben\u00f6tigen wir in erster Linie Zugang zur Live-Chatbot-Oberfl\u00e4che sowie ein Briefing zum Nutzungskontext (z. B. Zielsetzung, Zielgruppe, unterst\u00fctzte Sprachen). Interne Systemdokumentationen oder Konfigurationszug\u00e4nge sind f\u00fcr das Audit nicht erforderlich. Im Falle von Bots mit geschlossener Benutzergruppe ben\u00f6tigen wir zus\u00e4tzlich einen Testzugang.\n\nWenn Sie die Option \u201eUse Case Entwicklung\u201c nicht buchen, stellen wir Ihnen ein Formular zur Beschreibung Ihrer Use Cases zur Verf\u00fcgung. Falls Sie die Option gew\u00e4hlt haben, entwickeln wir geeignete Testszenarien auf Basis Ihres Briefings und stimmen diese vor Durchf\u00fchrung des Audits mit Ihnen ab.<\/p>\n<p data-start=\"4233\" data-end=\"4511\">Wenn Sie keine Use-Case Zusatzoption buchen, erhalten Sie von uns eine Ausf\u00fcllhilfe mit der Sie uns zu testende Use-Cases \u00fcbermitteln. W\u00e4hlen Sie die Zusatzoption Use-Case Entwicklung, entwickeln wir die Use-Cases und lassen diese vor Durchf\u00fchrung von Ihnen sichten.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_9 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Wie lang dauert ein Chatbot Audit?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p data-start=\"4562\" data-end=\"4793\">Das Level-1-Chatbot-Audit wird innerhalb von f\u00fcnf Arbeitstagen nach Erhalt des Briefings und \u2013 falls erforderlich \u2013 der Zugangsinformationen abgeschlossen.<\/p>\n<p data-start=\"4562\" data-end=\"4793\">Wenn Sie die Option \u201eUse Case Entwicklung\u201c buchen, planen Sie bitte zus\u00e4tzlich zwei Arbeitstage f\u00fcr die Erstellung der Testf\u00e4lle ein.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_10 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Wie sichern Sie die Vertraulichkeit?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p data-start=\"4562\" data-end=\"4793\">Alle Aktivit\u00e4ten und Ergebnisse werden grunds\u00e4tzlich vertraulich behandelt. Berichte und Ergebnisse werden nur mit den jeweiligen Kunden geteilt. Davon ausgenommen sind (anonymisierte) numerische Ergebnisse f\u00fcr unser Best-in-Class Benchmarking.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_11 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Welche Optionen k\u00f6nnen zus\u00e4tzlich gebucht werden?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p data-start=\"194\" data-end=\"373\">Das <strong data-start=\"198\" data-end=\"231\">9senses Level 1 Chatbot Audit<\/strong> kann an Ihre Bed\u00fcrfnisse angepasst werden. Zus\u00e4tzlich zur Basisversion stehen folgende Optionen zur Verf\u00fcgung:<\/p>\n<ul data-start=\"375\" data-end=\"1427\">\n<li data-start=\"375\" data-end=\"616\">\n<p data-start=\"377\" data-end=\"616\"><strong data-start=\"377\" data-end=\"398\">Use Case-Entwicklung<\/strong><br data-start=\"398\" data-end=\"401\" \/>In der Grundversion ben\u00f6tigen wir 3-5 relevante Gesch\u00e4ftsf\u00e4lle als Testszenarien durch Sie zur Verf\u00fcgung gestellt, abh\u00e4ngig von den Zielsetzungen Ihres Bots (z.B. Serviceabwicklung, Produktinformation, usw.). Falls Sie es bevorzugen w\u00fcrden, dass wir diese Testszenarien f\u00fcr Sie erarbeiten, w\u00e4hlen Sie bitte diese Zusatzoption.<\/p>\n<\/li>\n<li data-start=\"618\" data-end=\"842\">\n<p data-start=\"620\" data-end=\"842\"><strong data-start=\"620\" data-end=\"646\">Pr\u00fcfung von Bots mit offener Suche<\/strong><br data-start=\"646\" data-end=\"649\" \/>Falls Ihr Chatbot nicht nur Informationen von Ihrer eigenen Datenbasis bezieht, sondern auch auf Seiten von Drittanbietern oder im offenen Internet sucht, w\u00e4hlen Sie bitte diese Zusatzoption.<\/p>\n<\/li>\n<li data-start=\"844\" data-end=\"1058\">\n<p data-start=\"846\" data-end=\"1058\"><strong data-start=\"846\" data-end=\"883\">Bots mit Login-Voraussetzung<\/strong><br data-start=\"883\" data-end=\"886\" \/>Wenn Ihr Chatbot nur \u00fcber einen Zugang (Login) erreichbar ist, w\u00e4hlen Sie bitte diese Option. Wir ben\u00f6tigen in diesem Fall einen Testaccount auf Ihrem System.<\/p>\n<\/li>\n<li data-start=\"1060\" data-end=\"1245\">\n<p data-start=\"1062\" data-end=\"1245\"><strong data-start=\"1062\" data-end=\"1100\">Testen von Mehrsprachigkeit<\/strong><br data-start=\"1100\" data-end=\"1103\" \/>Wir \u00fcberpr\u00fcfen den Bot auf Mehrsprachigkeit, dazu geh\u00f6ren Aspekte wie Sprachwechsel, Konsistenz und \u00dcbersetzungsqualit\u00e4t (aktuell nur f\u00fcr gewisse Sprachen verf\u00fcgbar - siehe Zusatzoptionsauswahl).<\/p>\n<\/li>\n<li data-start=\"1247\" data-end=\"1427\">\n<p data-start=\"1249\" data-end=\"1427\"><strong data-start=\"1249\" data-end=\"1271\">Executive Briefing<\/strong><br data-start=\"1271\" data-end=\"1274\" \/>Buchen Sie eine 30-min\u00fctige managementorientierte Erl\u00e4uterung und Einordnung der Ergebnisse.<\/p>\n<\/li>\n<\/ul>\n<p data-start=\"1429\" data-end=\"1555\">Diese Optionen erm\u00f6glichen eine Anpassung an Architektur, Risikoprofil und Governance-Anforderungen.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_12 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Was ist der Unterschied zwischen Level 1 und Level 2?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Das Level 1 Audit basiert auf der externen Pr\u00fcfung des beobachtbaren Verhaltens w\u00e4hrend unserer Testszenarien aus Benutzersicht.<\/p>\n<p>Das Level 2 Audit basiert auf einer ma\u00dfgeschneiderten, vertieften Analyse, z.B. der technischen Architektur, Retrieval-Systematik, Governance, Compliance und Business Value.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_13 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">Wann sollte man ein Level 2 Audit durchf\u00fchren?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Ein Level 2 Audit ist dann sinnvoll, wenn der Chatbot in einem Level 1 Audit in den Kategorien Antwort- oder Dialogqualit\u00e4t einen geringen Wert von unter 3,5 erzielt. Mittels einer tiefergehenden Analyse k\u00f6nnen wir so die technischen Hintergr\u00fcnde konkret nachvollziehen und gezielte Handlungsoptionen aufzeigen.<\/p>\n<p>Ebenso ist ein Level 2 Audit sinnvoll, wenn der Bot in einem geschlossenen Nutzerkontext (z. B. f\u00fcr Kunden oder Mitarbeitende) eine zentrale Funktion mit entsprechendem Gesch\u00e4ftsrisiko \u00fcbernimmt.<\/p>\n<\/div><\/div>\n\n<div class=\"et_pb_accordion_item_14 et_pb_accordion_item et_pb_toggle et_pb_module et_pb_toggle_close et_flex_module preset--group--divi-accordion-item--divi-box-shadow--default preset--group--divi-accordion-item--divi-sizing--default\"><h3 class=\"et_pb_toggle_title\">K\u00f6nnen auch LLM-basierte (z.B. auf Basis ChatGPT) Chatbots gepr\u00fcft werden?<\/h3><div class=\"et_pb_toggle_content et_flex_module\"><p>Ja. Das 9senses Chatbot Audit ist auf regelbasierte Bots, Retrieval-Augmented-Generation-Systeme (RAG) sowie Large-Language-Model-basierte Assistenten anwendbar. Die Methodik konzentriert sich auf beobachtbare Leistung, Containment-Verhalten, Halluzinationsrisiken und Governance \u2013 nicht auf die technische Implementierung.<\/p>\n<\/div><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n<div class=\"et_pb_row_7 et_pb_row et-vb-row--no-module et_flex_row preset--group--divi-row--divi-box-shadow--default preset--group--divi-row--divi-sizing--default\">\n<div class=\"et_pb_column_14 et_pb_column et-last-child et_flex_column et_pb_column_empty et_pb_css_mix_blend_mode_passthrough et_flex_column_24_24 et_flex_column_24_24_tablet et_flex_column_24_24_phone et_flex_column_24_24_tabletWide preset--group--divi-column--divi-box-shadow--default preset--group--divi-column--divi-sizing--default\"><\/div>\n<\/div>\n<\/div>","protected":false},"excerpt":{"rendered":"","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_acf_changed":false,"footnotes":""},"class_list":["post-224422","page","type-page","status-publish","hentry"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.9senses.ai\/de\/wp-json\/wp\/v2\/pages\/224422","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.9senses.ai\/de\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.9senses.ai\/de\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.9senses.ai\/de\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.9senses.ai\/de\/wp-json\/wp\/v2\/comments?post=224422"}],"version-history":[{"count":0,"href":"https:\/\/www.9senses.ai\/de\/wp-json\/wp\/v2\/pages\/224422\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.9senses.ai\/de\/wp-json\/wp\/v2\/media?parent=224422"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}