Know what they want, before they ask.

30+ intent categories with confidence scores, sentiment classification per conversation, and urgency. At session end, every completed conversation also gets a summary. None of that requires extra configuration; it is the baseline.

The score is built in three layers: a heuristic score from 0 to 100, a signal breakdown your reps can read, and a written narrative explaining the number. The signals include buying language, engagement, urgency, and what Fred surfaced crossed with what the visitor did about it. Fred scores the conversation, not the browser URL, because when Fred is the navigation, the URL means nothing. The math is shown on every lead, not hidden behind a label.

Through a consent step, and Fred reads the conversation to time the ask rather than demanding an email at hello. Consent capture is GDPR and CCPA aligned, and the moment details are shared they are in your CRM in seconds.

Yes. High-priority intents like pricing inquiries, demo requests, or complaints can route to specific team members or workflows. The routing rules are yours, and escalation to a human always wins when the moment calls for it.

Yes. A frustrated visitor is a routing decision, not a sales opportunity. Negative sentiment biases Fred toward de-escalation and handoff, and the sentiment is recorded with the conversation so your team sees the temperature before they pick up the phone.

Yes. Stores that need product-level intent, like detecting when a shopper is comparing two specific items, can add the advanced e-commerce intent module. It is a separate capability for stores that want it, on top of the conversation intelligence every Fred already ships with.