Frequently Asked Questions
Getting Started
Q: How do I get started with TruthVouch? A: Sign up for a free account, take the AI Advisor assessment (free, no credit card), then explore the product based on your persona (CEO, CTO, Compliance Officer, Marketing Lead).
Q: Do I need technical knowledge to use TruthVouch? A: No. TruthVouch works for both technical and non-technical users. CTOs can use APIs and SDKs; business users can use dashboards and no-code tools.
Q: What if I don’t have existing AI systems to monitor? A: Start with the free AI Advisor assessment to understand your AI maturity, then implement recommendations step-by-step.
Security & Compliance
Q: Is my data secure? A: Yes. TruthVouch uses AES-256 encryption at rest, TLS 1.3 in transit, multi-tenant isolation with Row-Level Security, and annual third-party penetration testing. We’re SOC 2 Type II certified.
Q: Where is my data stored? A: Default: US (AWS us-east-1). You can choose EU (aws eu-west-1, GDPR compliant) or custom region on Enterprise.
Q: Are you GDPR compliant? A: Yes. We have a Data Processing Agreement, support all data subject rights, provide 72-hour breach notification, and maintain audit trails for 7 years.
Q: Do you have a SOC 2 report? A: Yes. SOC 2 Type II certified. Report available under NDA for Enterprise customers and serious prospects.
Q: Can I have TruthVouch on my own infrastructure? A: Yes. Self-hosted deployment available for Business and Enterprise tiers (Truth Firewall and Sentinel on-premises; dashboards cloud-hosted or self-hosted).
Features & Functionality
Q: How does hallucination detection work? A: TruthVouch queries monitored LLMs, extracts factual claims, and verifies against your truth nuggets using Natural Language Inference (NLI). 94%+ accuracy. Takes <200ms.
Q: What LLMs does TruthVouch monitor? A: ChatGPT, Claude, Gemini, Perplexity, Cohere, Mistral, Llama, Azure OpenAI, AWS Bedrock, and more. 9+ providers monitored simultaneously.
Q: Can I use TruthVouch with my own LLMs? A: Yes. Use the Governance Gateway SDK or REST API to integrate any LLM provider. Self-hosted Firewall also supported.
Q: How long does certification take? A: <1 minute for simple content. Complex documents with many claims: 1-5 minutes. Batch certification: 1-30 seconds per document.
Q: Can I integrate TruthVouch with my existing systems? A: Yes. APIs, SDKs (Python, TypeScript, .NET, Go, Java), webhooks, Slack/Teams, PagerDuty, Jira, ServiceNow, and more.
Q: Is there a free trial? A: AI Advisor assessment is completely free (no card needed). Platform trial available for 14 days (paid tier required).
Pricing & Billing
Q: What does each tier include? A: Starter ($349/mo): Hallucination monitoring, alerts, dashboard. Professional ($1,199): Includes Brand Intelligence, Advisor. Business ($2,499): Adds Governance, Certification, TruthOps. Enterprise: Custom pricing, everything unlocked.
Q: Can I upgrade or downgrade anytime? A: Yes. Changes take effect on next billing cycle. No penalties.
Q: Do you offer annual discounts? A: Yes. 15-20% discount for annual payment. Contact sales for custom pricing at scale.
Q: What about usage-based billing? A: All tiers are seat-based (unlimited API calls, monitoring). No per-request charges. Enterprise custom pricing available.
Integration & Development
Q: How do I integrate the SDK?
A: Install package (pip install truthvouch, npm install @truthvouch/sdk, etc.), initialize client with API key, start using. Takes 5 minutes.
Q: Are there code examples? A: Yes. Full documentation with examples in Python, TypeScript, C#, Go, Java. Plus integration guides for LangChain, LlamaIndex, Semantic Kernel, Spring AI.
Q: Can I use TruthVouch with RAG systems? A: Yes. Fact-check RAG responses before returning to users. Step-by-step guide available.
Q: How do webhooks work? A: TruthVouch sends HTTPS POST to your URL when events occur (new hallucination, certificate revoked, etc.). Payloads include digital signature for verification.
Support & Operations
Q: What’s the SLA? A: Professional: 99.5% uptime. Business: 99.9% uptime. Enterprise: 99.99% uptime with SLA credits.
Q: How do I get support? A: Email support@truthvouch.com, in-app chat (Monday-Friday 9am-5pm PT), community Slack (all tiers). Enterprise: phone support.
Q: Can I contact sales? A: Yes. sales@truthvouch.com or use the website chat. Demo available with Enterprise prospects.
Q: Do you have a status page? A: Yes. status.truthvouch.com shows real-time system status. Subscribe for alerts.
Advanced Topics
Q: How does the Neural Cache work? A: L1 (Redis) caches exact queries (<1ms). L2 (pgvector) caches semantically similar queries (5-10ms). Saves 80% of LLM calls, 95% cost reduction.
Q: What’s the governance pipeline? A: 15-stage processing inspecting every request/response: PII masking, injection detection, policy evaluation, fact-checking, etc. All <200ms.
Q: Can I write custom policies? A: Yes. Policies written in Rego (Open Policy Agent language). Full documentation and examples available.
Q: How do I migrate from manual monitoring? A: We have a migration assistant: audit current approach, build TruthVouch policies, run parallel for 30 days, then fully switch.
Still have questions? Email docs@truthvouch.com or start a chat.