Home Getting Started

Getting Started

Partha
By Partha
2 articles

What is Thunaivi?

Thunaivi is a 24/7 AI customer support platform that unifies WhatsApp, Instagram, Facebook, Telegram, and Website Chat in one inbox. AI handles repetitive queries instantly, and complex conversations are handed over to human agents with full context. Article 2: Quick Start (Go live in 30 minutes) 1. Create your workspace with official business email 2. Add company name, timezone, and support details 3. Connect at least one channel (start with Website Chat or WhatsApp) 4. Upload top FAQs/policies to AI Brain 5. Set handoff rules for sensitive topics 6. Run test conversations (inbound + outbound) 7. Go live and monitor first-day conversations Article 3: Minimum Setup Checklist Before Go-Live - Workspace created - Team members added - At least one channel connected - FAQs uploaded (pricing, refund, delivery, onboarding) - AI fallback + human handoff rules enabled - SLA notifications configured - Billing active (for paid plan users) Article 4: First Conversation Test Guide - Send a test message from connected channel - Confirm message appears in unified inbox - Check AI reply accuracy - Trigger a handoff case and confirm agent takeover - Validate response timestamps and assignment flow Article 5: Go-Live Readiness Score (Simple) - Channels ready: 30% - AI knowledge quality: 30% - Handoff safety: 20% - Team routing/SLA: 20% If total is below 85%, fix setup before public launch. Article 6: Common Mistakes in First Week - Launching without FAQ content - No escalation rules for sensitive queries - Missing ownership/assignment rules - Ignoring first 3 days of QA review - Using personal/free email for operations Article 7: Who should use Thunaivi first? Best starting teams: ecommerce, SaaS, education, healthcare, and lead-heavy businesses with repetitive support questions and slow first response times. Article 8: What success looks like in 14 days - Faster first response time - Reduced repetitive tickets - Better agent focus on complex cases - Clear channel-level visibility - Stable AI + human handoff quality

Last updated on Mar 05, 2026

Automation Rules & Routing

1) What are automation rules? Automation rules route, tag, assign, prioritize, and escalate conversations automatically based on channel, intent, message content, and SLA risk. 2) Core rule types to configure first - Intent-based routing (billing, onboarding, technical) - Channel-based routing (WhatsApp, Instagram, etc.) - Priority rules (urgent keywords/customers) - Escalation rules (risk/sentiment/compliance) - SLA warning rules (response delay) 3) Rule order strategy Use this order: 1. Block/restricted topics 2. Urgent priority rules 3. Intent routing rules 4. Assignment/queue rules 5. Follow-up reminders/SLA triggers Top-down rule order avoids conflicts. 4) Intent routing setup (recommended) Example: - If intent = billing → Queue: Billing Team - If intent = onboarding → Queue: Success Team - If intent = technical → Queue: Support L2 Always keep fallback queue: General Support. 5) Priority automation Set high priority when: - “refund”, “charged twice”, “legal”, “urgent”, “complaint” - VIP customer tag - Repeat contact within short period - No response after SLA threshold Then notify team lead automatically. 6) Assignment rules - Round-robin for normal queue - Skill-based for technical/billing - Sticky owner for ongoing conversation - Reassignment timeout if no first action in X minutes Keep ownership clear at all times. Track escalated vs resolved-without-escalation. 8) Rule testing checklist Before enabling: - Test with sample messages per intent - Validate label/priority assignment - Confirm queue/owner mapping - Confirm no loop/conflict - Check audit log event is recorded Enable rules only after test pass. 9) Common mistakes - Too many overlapping rules - No fallback queue - Missing urgent keyword set - No SLA triggers - Changing rules without version notes 10) Monthly optimization workflow - Review top escalations - Review misrouted conversations - Add new intent keywords - Remove unused rules - Compare KPI before/after rule changes

Last updated on Mar 05, 2026