Frequently Asked Questions

Our AI chatbot uses large language models trained on your business data. It understands natural language, maintains context across conversations, and provides relevant answers. The bot learns from your knowledge base, FAQs, product catalog to deliver accurate, brand-consistent responses.

Customer support (FAQs, ticket routing), sales automation (lead capture, qualification, follow-ups), e-commerce (cart recovery, order tracking, recommendations), marketing (campaigns, nurture sequences), and workflow automation (notifications, data sync, task creation).

Yes, true omnichannel capabilities. Deploy on website, mobile app, WhatsApp, Facebook Messenger, Instagram DMs, Telegram, SMS, email. All conversations unified in single dashboard with complete history — customers switch channels seamlessly without losing context.

Absolutely. Visual workflow builder lets you design custom paths, set conditional logic, create branching scenarios. Customize greetings, questions, responses, buttons, quick replies, and handoff rules. Pre-built templates available to modify for your needs.

50+ languages for customer conversations: English, Spanish, French, German, Portuguese, Italian, Dutch, Russian, Chinese, Japanese, Korean, Arabic, and more. AI automatically detects language and responds accordingly. Set preferences per channel or customer segment.

Automatic scoring based on: demographic fit, behavioral signals (pages viewed, time on site), engagement level, explicit qualifying answers, and BANT criteria. Scores calculated real-time, triggering actions like routing to sales or adding to nurture sequences.

Yes, essential for best results. Upload knowledge base, FAQs, product documentation, support tickets, website content. AI learns your terminology, products, policies, brand voice. Provide example conversations and corrections to continuously improve. Updates apply without downtime.

When AI can't help or detects frustration, seamless handoff to human agent. Agent receives full conversation context, customer history, AI-generated summary. Configure triggers (keywords, sentiment, topic) and routing rules (team, skill, availability). Customers can request human anytime.

Comprehensive built-in analytics: conversation volume, resolution rates, response times, CSAT, lead conversion, revenue attribution, handoff rates. Real-time dashboards, customizable reports, scheduled exports. Set up alerts and A/B test different approaches.

Full visual customization: colors, fonts, widget position, size, avatar, launcher icon, welcome message, animation styles. Use your logo, custom CSS, different appearances for different channels/pages. Embed as floating widget or inline component.

Subscription-based with monthly or annual billing (annual saves 20%). Plans based on conversation volume, features, team members. Set number of AI conversations/month, additional at per-unit rates. No hidden fees.

Yes! 14-day free trial with full feature access — no credit card required. Set up chatbot, test integrations, see real results. Then choose a plan or convert to limited free tier.

Starter: AI chatbot, 1 channel, basic analytics, email support. Growth: +3 channels, CRM, advanced analytics, live chat handoff. Pro: unlimited channels, all integrations, custom branding, priority support. Enterprise: dedicated instance, custom SLA, account manager.

Absolutely. Upgrade anytime — new features available immediately, prorated billing. Downgrade at billing cycle end. All data, configurations, history preserved. No lock-in contracts for standard plans.

Yes, custom pricing based on requirements. Features: dedicated infrastructure, 99.99% SLA, SSO/SAML, advanced security, unlimited team members, priority support, dedicated account manager, custom integrations, on-premise options.

Major credit/debit cards (Visa, Mastercard, Amex), PayPal for standard plans. Enterprise: bank transfer/wire, purchase orders with NET-30. All payments processed securely through Stripe.

Chatbot continues working — we never cut service. Additional conversations billed at competitive per-conversation rate. Alerts at 80% and 100% usage so you can upgrade. Enterprise has custom overage arrangements.

Yes! Nonprofits: 30% off. Early-stage startups (under $1M funding): 50% off first year via Startup Program. Educational institutions: special pricing. Contact sales with verification documentation.

Yes, 30-day money-back guarantee for new customers. Not satisfied within 30 days of paid subscription? Full refund, no questions asked. This is in addition to the free trial.

Most see 5-10x ROI in first year. E-commerce: $50K+ support savings + $100K+ cart recovery. B2B: 45% higher conversion, 3x sales productivity. ROI calculator provided during demo.

Industry-leading security: AES-256 encryption at rest, TLS 1.3 in transit, regular penetration testing, SOC 2 Type II compliance, 24/7 monitoring. AWS enterprise infrastructure. Never sell or share data.

Fully GDPR compliant. Data Processing Agreements provided. Support data subject requests (access, deletion, portability). Privacy by design. EU data residency options. Built-in consent management tools.

Default: AWS US (us-east-1). EU customers can choose European residency (Frankfurt). Enterprise: select specific regions — US, EU, UK, Australia, Asia-Pacific. Data never leaves chosen region without consent.

Only your authorized team members. Semios staff access only for support (with permission) or maintenance. All access logged and auditable. Principle of least privilege — employees access only what's necessary.

SOC 2 Type II certified. GDPR and CCPA compliant. HIPAA ready for healthcare. PCI-DSS compliant for payments. Audit reports available for enterprise customers on request.

You control retention. Configure how long conversations/data stored — 30 days to unlimited. Automatic deletion for inactive records. Account closure = permanent removal within 30 days. Enterprise: custom policies.

Your data trains only YOUR AI model — never shared with others or used for general models without consent. Each customer's AI is isolated, learning only from their specific data. Opt-out available.

Yes, SSO on Pro and Enterprise plans. SAML 2.0 support with Okta, Azure AD, Google Workspace, OneLogin, Auth0. 2FA available on all plans, can be enforced organization-wide.

99.9% for standard plans, 99.99% for Enterprise. Auto-failover, load balancing, multi-region redundancy. Real-time status at status.semios.ai. Enterprise SLAs include service credits for downtime.

Documented incident response plan. Affected customers notified within 72 hours (24 hours for enterprise). Detailed incident reports, remediation steps, preventive measures. Post-incident reviews to prevent recurrence.

No, and that's not our goal. AI handles 60-80% of repetitive inquiries, provides instant responses, scales 24/7. Human judgment, empathy, and complex problem-solving remain essential for escalations, sensitive issues, high-value conversations.

AI can misunderstand ambiguous requests; may "hallucinate" plausible but incorrect info; can't truly empathize; struggles with highly specialized knowledge without training; can be manipulated by adversarial prompts. We build safeguards for these.

Multiple safeguards: content filters, confidence thresholds triggering human review, guardrails preventing unauthorized promises, topic restrictions, response review and flagging. Regular monitoring catches and corrects issues.

Built-in: real-time monitoring dashboards, automated alerts for unusual patterns, human intervention at any point, all AI responses logged and auditable, regular reviews, clear escalation paths.

AI acknowledges uncertainty honestly, offers to connect with human, or suggests alternatives. Never fabricates answers when unsure. Configurable fallback responses and escalation paths ensure customers always get help.

Only what you authorize. Guardrails prevent unauthorized commitments. Configure what AI can and cannot promise. Sensitive actions (refunds, pricing changes) require human approval. Full control over AI authority levels.

We use diverse training data, regular bias audits, and implement fairness testing. AI trained on YOUR data reflects your approved content. Report concerns anytime — we take bias seriously and continuously improve.

Immediate human escalation available. Flag the conversation, review what went wrong, correct the AI's knowledge. We help you respond appropriately to the customer. Learn from every mistake to prevent recurrence.

We recommend transparency — it builds trust. Most customers accept AI when disclosed. Platform supports disclosure messages and AI identity. Some jurisdictions require disclosure. Being upfront prevents backlash.

Review conversation logs regularly, identify misunderstandings, update knowledge base, provide corrections. AI learns from feedback. Analytics show where improvements needed. Continuous optimization is key to success.