PREGUNTAS FRECUENTES

Respuestas claras y abiertas a todas tus dudas técnicas y comerciales

Configuración hablando, 8 esferas WebGL a 60 FPS, precisión RAG con citas, cuotas por visitante y resúmenes diarios por correo.

Setup, Conversational Configuration & Appearances

How do I set up the ChatNOV enterprise AI assistant? Do I need coding or flowcharts?

No. Skip rigid node trees and complex flow builders. Tell the ChatNOV Studio copilot your goals, knowledge sources, and placement rules in plain language, then paste a single script tag onto your website.

How many live appearances and color combinations are included?

Every plan includes all 8 unified live engines loaded on demand: Particle, Slime, Prism (3D Crystal), Nebula (WebGL Frosted Glass), Pulse, Helix, Avatar (AI Portrait / Custom SVG/WebP), and Emblem, paired with hundreds of curated color series.

Does the website-integrated AI assistant slow down Core Web Vitals or clash with CSS?

No. The ~70 KB core script loads asynchronously with defer, optional WebGL skins load strictly on demand, and the entire UI renders inside an isolated Shadow DOM root with zero external dependencies.

Can I test changes without affecting live website visitors?

Yes. Edits apply to an isolated Draft sandbox first. Publish immutable snapshots (v1.0, v1.1) when ready and roll back to any prior release in one click.

Knowledge Base (RAG), Accuracy & Anti-Hallucination

Which knowledge sources can ChatNOV index?

ChatNOV crawls single URLs, full sitemap.xml feeds, uploaded PDFs, DOCX, TXT, and Markdown files, plus direct Q&A entries added inside Studio.

Will the AI assistant hallucinate prices or unverified promises?

No. Answers are strictly grounded in your verified RAG chunks with clickable source citations. Unanswered questions are logged directly into your Studio Knowledge Gap Inbox for one-click training.

Is ChatNOV protected against prompt-injection attacks?

Yes. Server-side guardrails intercept instruction-override and system-prompt extraction attempts before they reach the underlying model.

Visitor Quotas, Anti-DDoS Bot Shield & Daily Email Digests

How do you stop a single visitor from draining my monthly quota?

Every project enforces a default limit of 20 replies per visitor per day across composite IP, signed HttpOnly/first-party token, and browser fingerprint signals. You can customize this limit anytime in Studio.

Do bot floods or fake DDoS requests consume my monthly message turns?

No. Requests must pass cryptographic client-proof headers (X-Chatnov-Guard), honeypot traps, origin checks, duplicate-spam filters, and sliding-window burst limiters before the AI model is ever invoked.

Do conversation summaries and reports arrive daily by email?

Yes. Automated daily summary digests are enabled by default at no extra turn cost, delivering conversation highlights and captured leads straight to your inbox.

Can I get an instant email transcript for every single conversation?

Yes. Enable "Instant Email per Conversation" in Studio to receive full transcripts right after each chat concludes. Each instant transcript email consumes 5 message turns from your monthly quota.

Unlimited Languages, Voice Mode & Lead Capture

Is the AI bot limited to 15 languages or does it support every language?

While our website and widget UI are localized in 15 languages, the ChatNOV assistant itself speaks and understands every known human language (including RTL scripts) plus constructed languages like Esperanto, Klingon, Huttese, and Sindarin.

How does live voice mode work?

Visitors tap the waveform icon to speak naturally. Real-time speech recognition streams their question with zero interruption during natural pauses while monochrome waveform bars oscillate smoothly.

Why does the 2-stage lead capture card convert better than long forms?

Stage 1 captures and emails the visitor work email immediately so zero leads are lost on drop-off; Stage 2 smoothly collects optional name and project context.

Subscription Plans, Evaluation Account, Data Privacy & Compliance

Can I try ChatNOV on my website for free before choosing a paid plan?

Yes. Open an evaluation workspace without a payment card—including 100 replies, 1 project, 5 knowledge sources, 10 crawled pages, and all 8 live appearances to test directly on your website.

Are monthly message, source, and crawl quotas per project or shared across my workspace?

Quotas are shared across all projects inside your workspace rather than isolated per project. For example, Starter’s 2,000 monthly replies and 60 crawled pages cover the combined traffic of your projects.

Are there per-seat agent fees or hidden setup charges?

Never. Starter, Growth, and Scale all run on transparent monthly workspace quotas with zero per-seat surcharges. View current tier rates in the Pricing section on our homepage.

Do I need to list every subdomain separately in Allowed Domains?

No. Adding your bare domain without www (e.g., yoursite.com) automatically covers www.yoursite.com and all subdomains. Adding www.yoursite.com restricts usage strictly to that exact host.

If I customize a skin like Slime and test another orb, will I lose my custom palette?

No. ChatNOV Studio stores independent per-skin color and avatar profiles so switching between appearances never overwrites your customized settings.

Are my uploaded documents used to train public AI models?

Never. Workspace documents and embeddings are logically isolated, encrypted with AES-256-GCM, and never sold or used to train third-party models under GDPR, CCPA/CPRA, and KVKK.

¿Necesitas una integración empresarial a medida?

Trabaja directamente con el equipo de ingeniería de ORZED para ERP, CRM, flujos de datos personalizados y arquitecturas de IA.

Contactar con ORZED