{ "company": "xpander.ai", "slug": "xpander-ai", "website": "https://xpander.ai", "audit_date": "2026-04-11", "overall_score": 63, "tier": "Human-Dependent", "tier_as_published": "H", "pillars": { "P1": { "name": "Signal Architecture", "score": 16, "max": 25 }, "P2": { "name": "Clarity Stack", "score": 15, "max": 25 }, "P3": { "name": "Trust Envelope", "score": 13, "max": 20 }, "P4": { "name": "Velocity Triggers", "score": 6, "max": 10 }, "P5": { "name": "Gravity Design", "score": 13, "max": 20 } }, "criteria": [ { "id": "P1-A", "pillar": "P1", "name": "Structured Data", "score": 0, "max": 5, "evidence": "No JSON-LD, microdata, or RDFa markup found on homepage or pricing page. Despite being an enterprise platform for AI agents, xpander.ai has zero structured semantic markup for machine consumption." }, { "id": "P1-B", "pillar": "P1", "name": "Machine-Readable Pricing", "score": 3, "max": 5, "evidence": "Pricing page contains specific numeric values in a clear HTML table: Cloud plan $485/month (5 users, 5 agents), Self-Hosted $6,300/month (50 users), per-seat pricing ($19 user, $49 builder, $29 agent), and LLM token tiers with per-million rates. No schema.org/Offer markup applied, but structure is explicit and parseable." }, { "id": "P1-C", "pillar": "P1", "name": "llms.txt / Agent Layer", "score": 5, "max": 5, "evidence": "/llms.txt confirmed present at xpander.ai/llms.txt (last updated 2026-04-01). Contains structured overview, core capabilities list, key page URLs (/product/, /security, /pricing, /company, free trial link), and strategic positioning guidance. Explicitly designed for LLM consumption." }, { "id": "P1-D", "pillar": "P1", "name": "API / MCP Availability", "score": 4, "max": 5, "evidence": "MCP API reference documented at /api-reference/mcp. Python SDK documented at /api-reference/sdk. Unified REST API at /user-guide/scale/unified-api supporting multiple frameworks (LangChain, CrewAI, Autogen, Google ADK, AWS Strands). A2A (agent-to-agent) protocol listed as a deployment surface. No public OpenAPI spec or agent card URL found, preventing a full score." }, { "id": "P1-E", "pillar": "P1", "name": "Discoverability (GEO)", "score": 4, "max": 5, "evidence": "robots.txt explicitly grants full site access to GPTBot, ChatGPT-User, Bingbot, and Googlebot. llms.txt with documentation index at docs.xpander.ai/llms.txt. Featured in VentureBeat and NVIDIA Developer Blog. Deliberately optimized for AI crawler discovery." }, { "id": "P2-A", "pillar": "P2", "name": "Offer Completeness", "score": 3, "max": 5, "evidence": "Cloud plan ($485/mo with specific inclusions) and Self-Hosted ($6,300/mo) pricing are clearly presented. But a \"Custom\" enterprise tier requires contact \u2014 creating a human gate for the highest-value tier. What the product does is precisely described. Pricing + capability not consolidated into a single machine-parseable artifact." }, { "id": "P2-B", "pillar": "P2", "name": "Scope & Limits", "score": 4, "max": 5, "evidence": "Explicit per-plan limits: users, builders, and agents/workflows per plan tier. LLM token pricing broken into three tiers with specific input/output rates per million tokens (Premium: $7.50 in/$40 out; Standard: $2.50/$10; Economy: $0.40/$2). Quantitative and specific." }, { "id": "P2-C", "pillar": "P2", "name": "Substitution & Fallback Rules", "score": 2, "max": 5, "evidence": "Multi-model LLM support (Azure OpenAI, Amazon Bedrock, Google Vertex, OpenAI, NVIDIA, Hugging Face) implies agent-level fallback capability but no explicit fallback routing rules or \"if primary model unavailable\" logic is documented in accessible pages." }, { "id": "P2-D", "pillar": "P2", "name": "Conditional Logic Transparency", "score": 2, "max": 5, "evidence": "Custom enterprise tier conditions undisclosed (\"contact sales\"). Privacy policy and terms pages are blocked in robots.txt (Disallow: /cookie-policy, /terms, /privacy-policy), making terms inaccessible to automated review. Some conditional pricing logic is therefore opaque." }, { "id": "P2-E", "pillar": "P2", "name": "Semantic Precision", "score": 4, "max": 5, "evidence": "Highly precise language throughout: \"2,000+ integrations,\" specific per-million token rates for three model tiers, \"14-day free trial,\" \"5 users, 5 builders, 5 agents/workflows\" in base plan. Uses technical terminology accurately (MCP, RBAC, Kubernetes, A2A, SCIM, SSO). Minimal vague claims." }, { "id": "P3-A", "pillar": "P3", "name": "Verifiable Performance", "score": 2, "max": 5, "evidence": "Enterprise Custom tier includes \"support SLA\" as a listed feature, but specific SLA terms are not published. Customer testimonials from named individuals (Director of Engineering at Axis Security, unnamed F200 companies). Featured in VentureBeat and NVIDIA Developer Blog. No public status page found. All performance claims are self-reported." }, { "id": "P3-B", "pillar": "P3", "name": "Scoped Permissions", "score": 4, "max": 5, "evidence": "Direct evidence of granular permission architecture: \"granular scope control to configure exactly what permissions your agent needs for each service,\" RBAC, \"scoped operations\" per specialized agent, \"permission boundaries enforced by IT,\" individual user account permissions. Enterprise-grade scoped permissions model clearly exists." }, { "id": "P3-C", "pillar": "P3", "name": "Audit Trail", "score": 4, "max": 5, "evidence": "Direct evidence: \"full audit logging for every action,\" \"Activity view provides full visibility into operations, API calls, and LLM responses,\" \"named agents with individual audit trails,\" \"every action logged with full context.\" Machine-accessible API for these logs not confirmed from available documentation." }, { "id": "P3-D", "pillar": "P3", "name": "Behavioral Consistency", "score": 3, "max": 5, "evidence": "Kubernetes-native with version-controlled infrastructure. \"No hidden cross service charges\" explicitly stated. SOC 2 alignment mentioned. Helm charts and versioned deployment implied. No published change-notice period or deprecation policy found, preventing a full score." }, { "id": "P4-A", "pillar": "P4", "name": "Friction-Free Activation", "score": 3, "max": 5, "evidence": "14-day free trial available, no credit card required, self-serve onboarding. \"Most teams have a working prototype in days.\" However, the minimum paid plan starts at $485/month \u2014 a significant price floor that represents a human approval gate for most agent use cases. Not truly frictionless for autonomous agent-initiated adoption." }, { "id": "P4-B", "pillar": "P4", "name": "Agent Decision Signals", "score": 3, "max": 5, "evidence": "Three model tiers (Premium/Standard/Economy) with explicit pricing provide agents with structured cost signals. Multiple framework compatibility options give agents clear decision criteria. But no programmatic \"here's when to upgrade\" signal, no usage-based threshold alerts via API, and no explicit agent-legible trial-expiry notification found." }, { "id": "P5-A", "pillar": "P5", "name": "Integration Depth / Switching Cost", "score": 4, "max": 5, "evidence": "2,000+ pre-built tool integrations. Deep enterprise system connections (Salesforce, Jira, GitHub, BigQuery, Confluence, ServiceNow, Snowflake). Custom agent containers with memory usage and CPU usage metrics in the enterprise tier. Kubernetes-native deployment with custom infrastructure creates very high switching cost once embedded." }, { "id": "P5-B", "pillar": "P5", "name": "Agent Memory / Personalization Layer", "score": 4, "max": 5, "evidence": "The platform is itself an agent memory and execution layer. \"AI handles mapping, Self-adapting\" workflows that adapt to API changes without manual field mapping. Personal AI assistants maintain organizational context across sessions. Agent memory is a core architectural component, not an add-on." }, { "id": "P5-C", "pillar": "P5", "name": "Programmatic Renewal Signals", "score": 2, "max": 5, "evidence": "Automated billing and subscription management exist. But no agent-accessible API for renewal status, usage-against-plan alerts, or programmatic subscription management found in audited documentation." }, { "id": "P5-D", "pillar": "P5", "name": "Compounding Value Signal", "score": 3, "max": 5, "evidence": "\"Self-adapting workflows\" that improve over time without manual field mapping. Platform builds knowledge of organizational data structures across deployments. But no agent-readable signal exposing the magnitude of compounding value (e.g., a \"workflow efficiency improvement\" metric or an agent-accessible \"how much better you've gotten\" API)." } ], "strongest_signals": [ { "title": "llms.txt Present and Structured", "detail": "xpander.ai has a current (2026-04-01), well-formed llms.txt file with overview, capabilities, key page URLs, and strategic positioning \u2014 one of the strongest agent-identity signals in the market. Combined with explicit AI crawler permissions in robots.txt, this is genuinely agent-forward infrastructure." }, { "title": "Deep Audit Logging + Scoped Permissions", "detail": "Direct evidence of full audit logs per action, individual agent audit trails, RBAC, and granular scope control. This is the gold standard for agent trust infrastructure \u2014 enterprises buying agent services need exactly this to delegate safely." }, { "title": "Platform IS the Agent Memory Layer", "detail": "Unlike tools that add memory as an afterthought, xpander.ai's architecture makes memory, context, and self-adaptation core primitives. This creates powerful compounding retention for agents that have been running for weeks or months." }, { "title": "Explicit AI Crawler Permissions", "detail": "Robots.txt explicitly names GPTBot and ChatGPT-User with full Allow \u2014 not an accident. xpander.ai is deliberately positioning itself to be discovered by AI systems." } ], "critical_gaps": [ { "title": "Zero Structured Data Markup", "detail": "Despite being an agent platform, the homepage and pricing pages have no schema.org markup. A company selling to AI agents should be the first to implement rich structured data \u2014 this is a significant credibility gap." }, { "title": "Terms and Privacy Pages Blocked by robots.txt", "detail": "Policy pages are explicitly disallowed from crawling, meaning agents cannot autonomously review terms of service, data handling, or legal conditions before committing to a purchase. This is an ironic gap for a company focused on agent governance." }, { "title": "No Public Status Page", "detail": "SLA commitments are hidden inside the enterprise \"Custom\" tier and not publicly published. An agent evaluating reliability has no third-party-verified uptime data to act on." }, { "title": "$485/Month Price Floor Creates a Human Gate", "detail": "While technically self-serve, the lowest plan costs $485/month \u2014 a spend level that requires human budget authorization in virtually every organization. This blocks autonomous agent adoption before it can start." } ], "priority_actions": [ { "action": "Add schema.org structured data to homepage and pricing page", "points_gain": 5, "pillar": "P1", "effort": "Low" }, { "action": "Unblock terms and privacy pages in robots.txt", "points_gain": 3, "pillar": "P2", "effort": "Low" }, { "action": "Publish a public status page", "points_gain": 3, "pillar": "P3", "effort": "Low" }, { "action": "Add a free/low-cost agent developer tier", "points_gain": 4, "pillar": "P4", "effort": "Medium" } ], "executive_summary": "xpander.ai scores 63/100 (Human-Dependent), the highest score in this audit series for a platform that is genuinely building agent infrastructure from the inside out. Its strongest pillars are Signal Architecture (led by a live, structured llms.txt) and Gravity Design (built-in agent memory, 2,000+ integrations, and self-adapting workflows). The critical irony: a company explicitly building AI agent infrastructure has zero schema.org structured data, blocks its own terms pages from AI crawlers, and prices its entry tier at $485/month \u2014 all of which prevent the very agents it serves from evaluating and purchasing autonomously. Closing the structured data gap (low effort, high impact) and publishing a public status page would immediately push this into Emerging territory.", "rubric_version": "v1-2026-04 (20 criteria, 100 raw points; P3-E Agent Registration added to rubric v2 in 2026-06, not scored in this audit)", "framework": "Agent Native Offers \u2014 The Agent Sale framework", "source_file": "2026-04-11 \u2014 xpander.ai \u2014 Agent Native Offer Audit.md", "rank": 11 }