Find answers to common questions about Click4Assistance, including our live chat software, AI-powered customer support platform, security, pricing, integrations, onboarding, and support.
Find answers to frequently asked questions about Click4Assistance AI Configuration & Knowledge Management, including AI Agent configuration, knowledge sources, prompts, training, testing, optimisation, governance, business rules, human escalation and AI administration.
Creating an AI Agent typically involves creating the AI Agent, adding knowledge sources, testing and refining responses, then publishing it to an Experience or communication channel. Once published, it can begin assisting customers while remaining available for continuous improvement.
AI Agents can use approved organisational knowledge from websites, PDF and Word documents, CSV files, FAQs, internal knowledge bases and business systems through secure APIs, helping deliver accurate and consistent responses.
No. AI Agents only use approved organisational knowledge managed by administrators and do not automatically learn from customer conversations, ensuring responses remain accurate, consistent and governed.
AI knowledge should be reviewed and updated whenever organisational information changes, including new products, policies, pricing, documentation or business processes, to maintain accurate and relevant responses.
Yes. Organisations can create multiple AI Agents for different departments or services, each with its own knowledge sources, configuration and behaviour.
Yes. Each AI Agent can be trained using different knowledge sources relevant to its role, helping improve response quality and simplify knowledge management.
Yes. AI Agents should be tested before publication to verify response accuracy, knowledge coverage, escalation behaviour, workflows and customer journeys before interacting with users.
Starting with a focused AI Agent enables faster deployment, simpler testing and continuous improvement before expanding into additional knowledge areas and use cases.
Yes. AI Agents handle routine enquiries while seamlessly transferring more complex conversations to human advisors, preserving conversation history for a smooth customer experience.
Yes. AI Agents can provide around-the-clock assistance by answering questions, collecting information, guiding customers and capturing enquiries outside normal business hours.
Yes. Subject to the capabilities of the selected language model, AI Agents can communicate in multiple languages, helping organisations support diverse audiences while using the same approved knowledge base. Multilingual responses should be tested as part of deployment.
AI performance can be measured using reporting and analytics, including conversation volumes, resolution rates, escalation rates, customer satisfaction, popular questions and knowledge gaps. These insights help organisations continuously improve AI performance.
Successful AI deployments begin with trusted knowledge, a focused use case and thorough testing. Organisations should regularly review conversations, update knowledge, provide clear human escalation paths and use reporting to support continuous improvement.
AI Agents learn from approved organisational knowledge provided by administrators rather than manually programmed questions and answers. This enables them to respond using natural language while remaining aligned with trusted business information.
Knowledge Sources are the approved repositories of information used by AI Agents when generating responses. These can include websites, help centres, PDF and Word documents, CSV files, policies, internal documentation and business systems accessed through secure APIs.
Yes. Training from an organisation's website is one of the quickest ways to deploy an AI Agent. Website content can be used to answer questions about products, services, policies, contact information and other approved topics without rewriting existing content.
For the best results, include factual and regularly maintained pages such as products and services, FAQs, pricing, support information, contact details, knowledge base articles, policies and user guides. High-quality content helps produce more accurate AI responses.
Avoid using pages containing outdated news, temporary promotions, duplicate or incomplete content, search results or conflicting information. Regularly reviewing website content helps maintain AI response quality.
Yes. PDF documents are commonly used to provide detailed product information, technical documentation, policy documents and user guides, allowing organisations to reuse existing documentation as AI knowledge sources.
Yes. Microsoft Word documents can be uploaded as knowledge sources, enabling organisations to use existing manuals, procedures and guidance documents. Well-structured documents help improve AI response quality.
Yes. CSV files can be used as AI knowledge sources where they contain structured information such as product catalogues, pricing, service lists or reference data. Well-organised spreadsheets help improve AI response quality.
Yes. Most organisations combine websites, documents and business systems to create a comprehensive AI knowledge base. Using multiple approved sources enables broader coverage while maintaining a consistent conversational experience.
The quality of information is generally more important than the quantity. Accurate, well-written and up-to-date content typically delivers better AI responses than large volumes of duplicated or outdated documentation.
Effective knowledge sources are accurate, current, clearly written, well organised, free from duplication, consistent with organisational policies and reviewed regularly to maintain AI accuracy.
Not necessarily. Many organisations achieve excellent results using existing documentation. However, improving clarity, consistency and completeness can help AI Agents provide more accurate and reliable responses.
Knowledge should be reviewed whenever significant organisational changes occur, including new products, service updates, policy changes, website redesigns, new documentation or regulatory requirements.
Yes. Each AI Agent can use its own approved knowledge library, allowing organisations to create specialist AI Agents for different audiences such as Customer Support, Sales, Technical Support, Broker Support or Employee Services.
Meet Arti, your fully trainable AI Agent. Use your website and documents to teach Arti how to deliver fast, accurate, and personalised customer support.
Successful AI training starts with trusted organisational knowledge. Remove duplicate or conflicting content, organise documentation logically, keep knowledge up to date, test AI responses regularly, expand knowledge gradually and monitor customer conversations to identify new training opportunities.
AI Knowledge Management is the process of organising, maintaining and governing the information used by an AI Agent. Well-managed knowledge helps ensure responses remain accurate, consistent and aligned with current organisational information.
Providing more information does not always improve AI performance. Well-structured, accurate and current content is typically far more valuable than large volumes of duplicated or outdated documentation, helping AI deliver more reliable responses.
Yes. Duplicate or conflicting information can reduce the consistency of AI responses. Maintaining a single authoritative version of important information helps improve response quality and simplifies ongoing knowledge management.
Documentation should be organised using clear headings, descriptive titles, short sections, consistent terminology, frequently asked questions and step-by-step guidance where appropriate. Well-structured content benefits both AI Agents and human users.
Yes. Every knowledge source should have a nominated owner responsible for reviewing and updating content. Assigning ownership helps ensure information remains accurate as products, services and business processes evolve.
Whenever policies, procedures or products change, the associated AI knowledge should be reviewed and updated promptly. Regular reviews help ensure AI continues to provide accurate and up-to-date information.
Yes. Organisations can create separate knowledge libraries for different business areas such as Sales, Customer Services, Technical Support, HR, Finance and IT Support. Specialist knowledge libraries improve response relevance and simplify ongoing maintenance.
AI performs best when using information that is accurate, clearly written, well maintained, logically structured and free from unnecessary duplication. Frequently updated operational content generally delivers better results than marketing-focused material.
Yes. AI Agents can use approved internal documentation to help employees quickly access policies, procedures, technical guidance and operational information. Access can be controlled through appropriate authentication and user permissions.
Whenever new knowledge is added, organisations should test representative customer questions before publication. Testing should verify response accuracy, tone of voice, references, escalation behaviour and consistency across similar enquiries.
Yes. Knowledge should be reviewed as part of normal business operations rather than only when issues arise. Regular reviews help ensure content remains accurate, relevant and aligned with organisational objectives.
Yes. Reporting and conversation analysis can identify questions that AI cannot answer confidently or consistently. These insights help organisations expand their knowledge base and continuously improve AI performance.
Consistent AI responses are achieved through well-managed knowledge sources, standard terminology, regular content reviews, the removal of conflicting information, clear governance processes and ongoing testing.
Yes. AI knowledge can be organised so different departments, websites or AI Agents access only the information relevant to their responsibilities. This enables organisations to deliver specialised conversational experiences while managing knowledge from a single platform.
Successful knowledge management includes assigning ownership for every knowledge source, reviewing content regularly, removing duplicate information, maintaining a consistent writing style, testing new knowledge before publication, monitoring customer conversations and continuously improving the knowledge base over time.
Testing helps ensure AI Agents provide accurate, relevant and appropriate responses before they are made available to customers. It also identifies knowledge gaps, ambiguous answers and opportunities for improvement while reducing the risk of incorrect information being presented.
Testing should assess response accuracy, knowledge coverage, escalation behaviour, tone of voice, workflow integration, business rules and the overall customer journey to ensure the AI performs as expected.
Organisations should prepare representative questions covering FAQs, product enquiries, business processes, policies, complex scenarios, unusual wording and follow-up questions to evaluate AI performance in realistic customer interactions.
Yes. Involving teams such as Customer Services, Sales, Technical Support, Marketing, Compliance and Product Specialists helps identify knowledge gaps and opportunities for improvement that may not be found during technical testing alone.
If the AI Agent cannot confidently answer a question, it can ask the customer to rephrase the enquiry, offer related information, transfer the conversation to a human advisor, provide alternative contact methods or capture enquiry details for follow-up.
No. Many enquiries can be resolved by asking additional questions. However, conversations involving complex decisions, sensitive matters or requests beyond the AI Agent's approved knowledge should be escalated to an appropriate human advisor.
AI conversations should be reviewed regularly to identify new customer questions, missing knowledge, response improvements, process changes and emerging trends. Regular reviews support continuous improvement and help keep AI aligned with evolving business requirements.
Yes. Reporting provides valuable insight into how customers interact with AI Agents, including frequently asked questions, escalation rates, customer satisfaction, conversation volumes, unanswered topics and common search themes. These insights help organisations refine knowledge and continuously improve AI performance.
AI Governance is the framework of policies, processes and controls used to ensure AI operates in a responsible, secure and predictable manner. Effective governance includes approved knowledge sources, defined ownership, regular reviews, human oversight, change management and ongoing monitoring.
No. AI is most effective when it complements human expertise. Routine and repetitive enquiries can be automated, while conversations requiring judgement, empathy or customer-specific decisions should be transferred to appropriately trained human advisors.
AI success should be measured against business objectives using metrics such as customer satisfaction, resolution rates, reduced response times, lower contact centre demand, increased advisor productivity, greater self-service adoption and improved operational efficiency.
Yes. As organisations review conversations, update knowledge and refine customer journeys, AI performance can improve significantly. AI should be treated as a continuously evolving service rather than a one-time implementation project.
AI guardrails are the policies and controls that define how an AI Agent should operate. They include using approved knowledge sources, following organisational policies, protecting sensitive information, escalating when appropriate and maintaining consistent behaviour.
Human oversight enables organisations to monitor AI performance, review customer interactions and make improvements where required. It also helps ensure AI continues to reflect current business policies while keeping important decisions under appropriate human control.
Successful AI Governance includes defining clear ownership, testing thoroughly before deployment, monitoring conversations regularly, keeping knowledge up to date, providing appropriate human escalation, measuring performance against business objectives and continuously improving the customer experience. Managing AI as an ongoing business capability helps ensure it continues to deliver long-term value.
Customise every aspect of your chat window and widget from colours to icons for a smooth on-brand customer journey.
