What if your bot could tell the difference? Between a visitor who just needs a quick answer, someone actively comparing options, and a frustrated customer one click away from leaving?
Most bots can't. They deliver the same response to everyone, missing the nuance that separates a casual question from an urgent problem or a buying signal.
But here's what's possible now: With AI that understands sentiment, intent, and context, your bot can actually recognize what each customer needs in real-time. Upset customers get immediate attention. Ready-to-buy visitors get routed to sales reps. Explorers get guided through their options without friction.
Enter Criteria router + AI filters
If you’re using SalesIQ’s Criteria router card in the Codeless bot builder, you already know the potential of routing chats based on conditions you define. Now, we’ve made it smarter with AI Filters - Sentiment, Intent, and Topic.
- Sentiment: Understands the tone of the visitor (positive, negative, neutral). You can split the conversation based on the sentiment. Example: If a visitor sounds frustrated, their sentiment is marked negative, you can route them directly to experienced agents.
- Intent: Detects the purpose of the conversation (like cancellation, purchase, or scheduling). You can mention the intent and split the chat that matches your intent. Example: If someone says, “I’m not happy with my order”, the intent could be cancellation, you can trigger a survey or escalation.
- Topic: Here you can define topics like, iPhone in ecommerce, 2bhk flat in Chicago for real estate, basically nouns. If the bot finds these topics, it will route accordingly.
Webinar alert: Want to see more examples and deep-dive into use cases? Check out our AI-Powered engagement with SalesIQ webinar, where we covered all the new AI-based solutions in SalesIQ, including Criteria Router and AI Filters.
Let’s look at a few use cases across industries:
1. E-commerce: A customer types, “My package arrived broken. I want a refund.”
- Sentiment: Negative
- Intent: Cancellation/Return
Route this chat straight to your experienced support agents, skipping the bot’s small talk.
2. Real Estate: A visitor says, “Looking for a flat in Chicago near the lake.”
- Topic: Flat, Chicago
- Intent: Search
Instantly connect them with the right sales rep covering that location, instead of bouncing them around.
3. IT / Tech Support: Someone writes, “The app keeps crashing whenever I try to upload a file.”
- Sentiment: Negative
- Intent: Technical Issue
- Topic : App
Route to your technical app support team, not the general help queue.
Implementation tips:
1. Use context variables: AI filters works only with bot context variables. Make sure to:
- Use the text input card to capture customer messages
- Enable "Save in bot context"
- Give a variable name
- Use the same in your Criteria router card
2. Prioritize your rules: When multiple rules match, the first one wins. Structure your rules by priority: Example:
- Rule 1: Sentiment: Negative + Intent: Cancellation
- Rule 2: Sentiment: Negative + Topic: App issues
If a customer says "The app is crashing again, I want to cancel," both rules technically match, but Rule 1 triggers because it's listed first.
Get started today
AI Filters are live in your Criteria Router.
With this, your bot is no longer just a gatekeeper, it’s a smart filter. It understands urgency, emotion, and intent, and directs conversations to the right place, faster. That means happier customers, less frustration, and more chances to convert.
Set up your rules, test a few conversations, and watch how much smoother your chat flow becomes and let us know if how it works.