{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"tool_cloud-humans","slug":"cloud-humans","name":"Cloud Humans","type":"product","url":"https://en.cloudhumans.com","page_url":"https://unfragile.ai/cloud-humans","categories":["chatbots-assistants"],"tags":[],"pricing":{"model":"freemium","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"tool_cloud-humans__cap_0","uri":"capability://planning.reasoning.intent.based.query.classification.and.routing","name":"intent-based query classification and routing","description":"Cloud Humans implements a multi-stage classification pipeline that analyzes incoming customer queries to determine whether they can be resolved by AI or require human escalation. The system likely uses NLP-based intent detection (possibly transformer-based embeddings or rule-based classifiers) to categorize queries into predefined support categories, then applies confidence thresholds to decide routing. Queries below confidence thresholds or matching complex intent patterns are automatically routed to human agents, while high-confidence routine queries are handled by the AI layer.","intents":["I want to automatically handle simple FAQ-style questions without human involvement","I need to identify which customer issues are too complex for AI and require immediate human attention","I want to reduce the volume of tickets reaching my support team by filtering routine requests first","I need to understand what percentage of incoming queries my AI can confidently resolve"],"best_for":["SaaS companies with 20-40% repetitive support tickets","E-commerce platforms handling high-volume order status and FAQ inquiries","Support teams wanting to preserve human capacity for complex issues"],"limitations":["No transparency provided on classification accuracy rates or false-positive escalation costs","Confidence thresholds appear to be platform-managed, not customizable per client","Likely struggles with nuanced or context-dependent queries that don't fit predefined categories","No documented support for domain-specific intent taxonomies or custom classification models"],"requires":["Integration endpoint or API key for Cloud Humans platform","Incoming query data in text format (chat, email, or form submission)","Predefined support categories or FAQ knowledge base to train classification"],"input_types":["text (customer query)","structured metadata (customer ID, account type, issue category)"],"output_types":["routing decision (AI-handle vs human-escalate)","confidence score","suggested category or intent label"],"categories":["planning-reasoning","customer-service-automation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_cloud-humans__cap_1","uri":"capability://text.generation.language.ai.powered.conversational.response.generation.for.routine.inquiries","name":"ai-powered conversational response generation for routine inquiries","description":"Cloud Humans generates contextually appropriate responses to customer queries using a language model backend (likely GPT-based or similar), constrained by a knowledge base or FAQ database to ensure accuracy and brand consistency. The system likely implements prompt engineering with context injection (customer history, account details, relevant documentation) to produce personalized responses. Response generation is gated by the classification layer—only queries deemed routine and high-confidence trigger this capability, reducing hallucination risk and support costs.","intents":["I want to provide instant responses to common customer questions without human agent involvement","I need to maintain consistent, on-brand messaging across all AI-generated support responses","I want to reduce response time for FAQ-style inquiries from hours to seconds","I need to personalize responses based on customer account history or previous interactions"],"best_for":["Companies with well-documented FAQ or knowledge base content","Support teams handling high volumes of repetitive inquiries (order status, billing, password resets)","Businesses where response consistency and brand voice are critical"],"limitations":["Requires high-quality, up-to-date knowledge base or FAQ content to avoid hallucinated responses","No documented support for multi-turn conversations or complex problem-solving workflows","Response generation quality depends entirely on classification accuracy—misclassified queries may receive irrelevant answers","No transparency on how customer context (history, account data) is injected into prompts","Likely cannot handle edge cases or novel customer scenarios outside training data"],"requires":["Knowledge base or FAQ database with documented answers","Integration with customer data system (CRM, account management) for context injection","LLM API access (likely OpenAI, Anthropic, or proprietary model)","Predefined response templates or guardrails to constrain generation"],"input_types":["text (customer query)","structured customer context (account ID, purchase history, previous interactions)"],"output_types":["natural language response (text)","confidence score or quality metric"],"categories":["text-generation-language","customer-service-automation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_cloud-humans__cap_2","uri":"capability://automation.workflow.seamless.ai.to.human.agent.handoff.with.context.preservation","name":"seamless ai-to-human agent handoff with context preservation","description":"When a query is classified as requiring human intervention, Cloud Humans implements a handoff mechanism that transfers the conversation context (query history, customer metadata, classification reasoning) to a human agent without requiring the customer to re-explain their issue. The system likely maintains a conversation state object that includes the original query, any AI-generated analysis, customer account details, and escalation reason. Human agents access this context through a unified dashboard, enabling them to pick up the conversation mid-stream without context loss.","intents":["I want to escalate complex queries to human agents without losing conversation history","I need my support team to have full context (customer history, previous attempts) when they take over from AI","I want to prevent customers from having to repeat themselves when transferred from AI to human","I need to track why queries were escalated and what AI attempted before handoff"],"best_for":["Support teams using multiple tools (chat, email, ticketing) that need unified context","Companies where customer frustration from re-explaining issues is a known pain point","Mid-market SaaS with 20-40% escalation rates from AI to human"],"limitations":["Handoff latency and context transfer speed not documented","No transparency on how context is stored or secured during transfer","Unclear whether handoff works across multiple channels (chat to email, chat to phone)","No documented support for custom context fields or business-specific metadata","Potential for context bloat if AI generates verbose analysis that clutters human agent view"],"requires":["Active conversation session with customer","Human agent availability and dashboard access","Persistent session storage (likely database or message queue)","Integration with human agent interface (web dashboard or CRM)"],"input_types":["conversation history (text)","customer metadata (structured)","escalation reason (text or enum)"],"output_types":["context-enriched ticket or conversation object","human agent notification","conversation state snapshot"],"categories":["automation-workflow","customer-service-automation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_cloud-humans__cap_3","uri":"capability://data.processing.analysis.support.queue.volume.reduction.through.ai.deflection","name":"support queue volume reduction through ai deflection","description":"Cloud Humans measures and reports on the volume of queries successfully handled by AI versus those escalated to humans, providing visibility into deflection rates and support cost savings. The system tracks metrics like queries-per-hour handled by AI, escalation rate, average resolution time, and estimated human agent hours saved. This capability likely includes a dashboard or reporting interface that aggregates these metrics over time, enabling support managers to understand the impact of AI automation on their support operations and justify continued investment.","intents":["I want to measure how much support load the AI is actually removing from my team","I need to calculate ROI on the Cloud Humans platform based on human agent hours saved","I want to track deflection rates over time to understand if AI performance is improving","I need to identify which types of queries are being successfully automated vs escalated"],"best_for":["Support managers and operations leaders evaluating AI ROI","Finance teams calculating cost savings from automation","Companies with mature support operations that can benchmark against historical data"],"limitations":["No transparency on how deflection metrics are calculated or what constitutes a 'successful' resolution","Unclear whether metrics account for customer satisfaction or only volume reduction","No documented support for custom KPI definitions or business-specific metrics","Likely lacks granular attribution (which queries were deflected vs which were escalated due to misclassification)","No visibility into false-positive escalations that waste human agent time"],"requires":["Minimum baseline of queries processed (likely 100+ per day for meaningful metrics)","Historical support data for benchmarking (optional but recommended)","Access to Cloud Humans analytics dashboard or reporting API"],"input_types":["query routing decisions (AI-handle vs escalate)","resolution outcomes (resolved vs escalated)","timestamp and metadata"],"output_types":["deflection rate (percentage)","queries handled by AI (count)","estimated cost savings (currency)","escalation rate (percentage)","time-series metrics dashboard"],"categories":["data-processing-analysis","customer-service-automation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_cloud-humans__cap_4","uri":"capability://automation.workflow.freemium.tier.with.zero.commitment.onboarding","name":"freemium tier with zero-commitment onboarding","description":"Cloud Humans offers a freemium pricing model that allows customers to test the platform without providing payment information upfront, reducing friction for initial adoption. The free tier likely includes limited query volume (e.g., 100-500 queries/month) and basic features (intent classification, simple response generation, basic escalation). Customers can evaluate platform performance, integration complexity, and support quality before committing to paid plans, reducing perceived risk and enabling data-driven purchasing decisions.","intents":["I want to test Cloud Humans with my actual support data before committing budget","I need to evaluate whether the AI actually reduces our support load before paying","I want to understand integration complexity and time-to-value without upfront investment","I need to pilot the platform with a small subset of queries to validate ROI"],"best_for":["Mid-market SaaS and e-commerce companies with budget approval processes","Support teams skeptical of AI automation claims and wanting proof","Organizations with low risk tolerance for new vendor commitments"],"limitations":["Free tier query limits may be too restrictive to generate meaningful deflection metrics","No documented support for extending free tier or rolling free usage into paid plans","Unclear whether free tier includes human escalation or only AI response generation","Limited features in free tier may not reflect paid tier performance or capabilities","No transparency on free tier SLA or support quality vs paid tiers"],"requires":["Email address for account creation (no credit card required)","Integration endpoint or API key (likely provided immediately upon signup)"],"input_types":["customer email","basic account information"],"output_types":["API credentials","access to freemium dashboard","limited query quota"],"categories":["automation-workflow","customer-service-automation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_cloud-humans__cap_5","uri":"capability://tool.use.integration.multi.channel.query.ingestion.and.normalization","name":"multi-channel query ingestion and normalization","description":"Cloud Humans accepts customer queries from multiple input channels (chat, email, web forms, potentially SMS or social media) and normalizes them into a unified format for processing by the classification and response generation layers. The system likely implements channel-specific adapters that extract query text, customer metadata, and channel context, then map them to a canonical query object. This abstraction enables the AI and routing logic to operate independently of the source channel, while preserving channel-specific context (e.g., email subject line, chat session ID) for escalation and context preservation.","intents":["I want to handle customer queries from multiple channels (chat, email, forms) with a single AI system","I need to preserve channel-specific context (email thread, chat history) when escalating to human agents","I want to avoid building separate AI systems for each support channel","I need to normalize customer identity across channels so AI has full context"],"best_for":["Companies with omnichannel support operations (chat, email, web forms)","Support teams wanting to consolidate multiple chatbot and automation tools","Businesses where customer context spans multiple channels"],"limitations":["No documentation on which channels are supported (likely limited to web-based channels)","Unclear whether phone or voice channels are supported","No transparency on how customer identity is matched across channels","Potential latency in email channel due to async nature vs real-time chat","No documented support for custom channel integrations or webhooks"],"requires":["Integration with each supported channel (API, webhook, or embedded widget)","Customer identity system or CRM for cross-channel context","Channel-specific credentials (e.g., email account access, chat platform API key)"],"input_types":["text (query)","channel metadata (source, timestamp, customer ID)","channel-specific context (email subject, chat session ID)"],"output_types":["normalized query object","routing decision","channel-specific response (formatted for email, chat, etc.)"],"categories":["tool-use-integration","customer-service-automation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_cloud-humans__cap_6","uri":"capability://automation.workflow.human.agent.availability.and.capacity.management","name":"human agent availability and capacity management","description":"Cloud Humans manages the availability and workload of human agents, routing escalated queries to available agents based on capacity, skill level, or specialization. The system likely maintains an agent status model (available, busy, offline) and implements a queue or load-balancing mechanism to distribute escalated queries fairly. This capability may include features like agent skill tagging (e.g., 'billing', 'technical', 'account management') to route queries to specialists, and queue management to prevent agent overload or customer wait times.","intents":["I want to route escalated queries to available human agents without manual assignment","I need to prevent agent overload by distributing queries based on current workload","I want to route specialized queries (billing, technical) to agents with relevant expertise","I need to track agent availability and queue depth in real-time"],"best_for":["Support teams with multiple agents and varying skill levels","Companies with specialized support domains (billing, technical, account management)","Operations with variable query volume and need for load balancing"],"limitations":["No documentation on queue management algorithm or fairness guarantees","Unclear whether skill-based routing is automatic or requires manual configuration","No transparency on how agent availability is tracked or updated","Potential for queue buildup if escalation rate exceeds agent capacity","No documented support for agent scheduling or shift management"],"requires":["Human agents with Cloud Humans dashboard access","Agent status tracking system (likely integrated with dashboard)","Optional: skill or specialization tags for agents"],"input_types":["escalated query","agent availability status","optional: query specialization or skill requirement"],"output_types":["agent assignment","queue position","estimated wait time"],"categories":["automation-workflow","customer-service-automation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_cloud-humans__cap_7","uri":"capability://memory.knowledge.knowledge.base.integration.and.context.injection","name":"knowledge base integration and context injection","description":"Cloud Humans integrates with customer knowledge bases, FAQs, or documentation systems to ground AI response generation and improve classification accuracy. The system likely implements a retrieval mechanism (semantic search or keyword matching) that fetches relevant documentation snippets based on the customer query, then injects this context into the LLM prompt. This enables the AI to generate responses that align with documented support policies and reduces hallucination by constraining generation to verified information.","intents":["I want the AI to answer questions based on my actual documentation, not make things up","I need to ensure AI responses align with my support policies and brand voice","I want to automatically update AI responses when I update my knowledge base","I need to improve AI accuracy by grounding responses in verified information"],"best_for":["Companies with well-maintained knowledge bases or FAQ systems","Support teams where response accuracy and consistency are critical","Organizations with complex products requiring detailed documentation"],"limitations":["Requires high-quality, up-to-date knowledge base content—outdated docs lead to stale responses","No transparency on retrieval mechanism (semantic search vs keyword matching)","Unclear how context injection handles conflicting or ambiguous documentation","No documented support for knowledge base versioning or change tracking","Potential latency from retrieval step before response generation"],"requires":["Existing knowledge base or FAQ system (internal or third-party like Zendesk, Confluence)","Integration API or webhook to sync knowledge base with Cloud Humans","Structured or semi-structured documentation format"],"input_types":["customer query (text)","knowledge base documents (text, markdown, HTML)"],"output_types":["retrieved context snippets","grounded AI response","source attribution (optional)"],"categories":["memory-knowledge","customer-service-automation"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":41,"verified":false,"data_access_risk":"high","permissions":["Integration endpoint or API key for Cloud Humans platform","Incoming query data in text format (chat, email, or form submission)","Predefined support categories or FAQ knowledge base to train classification","Knowledge base or FAQ database with documented answers","Integration with customer data system (CRM, account management) for context injection","LLM API access (likely OpenAI, Anthropic, or proprietary model)","Predefined response templates or guardrails to constrain generation","Active conversation session with customer","Human agent availability and dashboard access","Persistent session storage (likely database or message queue)"],"failure_modes":["No transparency provided on classification accuracy rates or false-positive escalation costs","Confidence thresholds appear to be platform-managed, not customizable per client","Likely struggles with nuanced or context-dependent queries that don't fit predefined categories","No documented support for domain-specific intent taxonomies or custom classification models","Requires high-quality, up-to-date knowledge base or FAQ content to avoid hallucinated responses","No documented support for multi-turn conversations or complex problem-solving workflows","Response generation quality depends entirely on classification accuracy—misclassified queries may receive irrelevant answers","No transparency on how customer context (history, account data) is injected into prompts","Likely cannot handle edge cases or novel customer scenarios outside training data","Handoff latency and context transfer speed not documented","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.36666666666666664,"quality":0.7300000000000001,"ecosystem":0.15000000000000002,"match_graph":0.25,"freshness":0.75,"weights":{"adoption":0.25,"quality":0.25,"ecosystem":0.1,"match_graph":0.35,"freshness":0.05}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-05-24T12:16:29.717Z","last_scraped_at":"2026-04-05T13:23:42.552Z","last_commit":null},"community":{"stars":null,"forks":null,"weekly_downloads":null,"model_downloads":null,"model_likes":null}},"distribution":{"claim_url":"https://unfragile.ai/submit?claim=cloud-humans","compare_url":"https://unfragile.ai/compare?artifact=cloud-humans"}},"signature":"Z/vOKX1dtl/XHEJHpxVmARJrvw0mhNmU7GPN7wOAO5gGpT+M5EALoKUalqiJjph79UhMO/AIwU6C2j1yj5g/Bg==","signedAt":"2026-06-22T18:32:42.430Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/cloud-humans","artifact":"https://unfragile.ai/cloud-humans","verify":"https://unfragile.ai/api/v1/verify?slug=cloud-humans","publicKey":"https://unfragile.ai/api/v1/trust-passport-public-key","spec":"https://unfragile.ai/trust","schema":"https://unfragile.ai/schema.json","docs":"https://unfragile.ai/docs"}}