Your clients have a question. Your team is busy. The AI Assistant responds in 3 seconds.

Yes. It's a single agent deployed in parallel across up to 10 channels simultaneously: WhatsApp, website, phone (AI voice), email, Slack, Teams, SMS, Instagram, Facebook Messenger, and chatbot widget. Same configuration, same knowledge base, same personality across all channels. Conversational memory is shared: if a client starts by email then calls, the agent already knows the context. Think of a representative who can answer 1,000 calls, 500 WhatsApps, and 200 emails simultaneously, without ever getting tired or giving contradictory answers.

SkaLean implements a configurable confidence threshold system. Each response receives a confidence score (0-100%). If the score is below the threshold you define (e.g., 70%), the agent doesn't "make things up" — it automatically transfers to a human agent (Human-in-the-Loop) or explicitly says "I'm not certain, let me connect you with a specialist." The anti-hallucination protection limits responses to your knowledge base — essential for medical or legal sectors.

The PII module automatically detects and pseudonymizes 15 categories of sensitive data before any LLM processing: names, addresses, phone numbers, emails, card numbers, insurance numbers, dates of birth, file numbers, and 7 other types. Example: "My name is Marie Dupont, my Visa card ends in 4521" becomes "[NAME], my [PAYMENT] card ends in [FRAGMENT]" before reaching the LLM. The LLM never sees the real data. Natively HIPAA and GDPR compliant architecture.

AI voice uses state-of-the-art speech synthesis (TTS) and recognition (STT), with a response time of under 1 second between the end of a customer sentence and the beginning of the agent's response. Phone integration is done via SIP/WebRTC, compatible with all modern IP-PBX systems and Twilio. Real example: a dental clinic uses the voice agent to confirm appointments and answer pricing questions — 80% of incoming calls handled without a human, 7 days/week from 8am to 10pm.

1 credit equals approximately 500 tokens. A typical conversational exchange (question + response) consumes 2 to 4 credits, or $0.01 to $0.02. For 500 interactions/month with 4 exchanges each: 500 × 4 × 3 credits = 6,000 credits/month — covered by the Starter plan which includes 10,000 credits/month. Voice costs 10 credits/minute (≈$0.05/min). You are notified at 80% consumption to avoid any surprises.

Yes. Persistent memory retains context between sessions (identified by phone number, email, or WhatsApp ID). When a customer returns, the agent can access their interaction history, preferences, and open issues. Example: a customer who contacted your service about a billing issue last month calls back today — the agent can say "Hello, I can see you contacted our service on March 15th regarding invoice #4821. Has your issue been resolved?" Without the customer needing to repeat their context.

Completely. You configure: (1) the agent's name (e.g., "Sophie from Beauregard Clinic"), (2) tone (formal, warm, professional), (3) allowed and prohibited topics, (4) pre-validated responses for sensitive questions (pricing, warranties, contraindications), (5) languages and automatic escalations — if a customer mentions "lawyer" or "emergency," immediate transfer to a human. The agent operates under your visual identity: your customers never see "SkaLean."