Every conversation,
classified on the fly.
Define your own taxonomy — groups and categories — and let the AI tag every message instantly. Filter in Databot Live by topic, measure volume in Analytics and discover what your team really handles.
Nested groups and categories
Structure your vocabulary as a tree: topic groups with categories inside. Drag to reorder.
AI tags in real time
Every message is classified the moment it arrives. No queues, no nightly batch jobs, no humans tagging by hand.
Live filters in Databot Live
Operators see the queue filtered by category, and Analytics reports volume and trends by topic and channel.
You design the taxonomy. AI does the rest.
Groups, categories, order — your way.
Create topic groups (Products, Billing, After-sales…) and inside each one the fine categories the AI will use as labels. Reorder with drag & drop, describe each category with examples and synonyms, version your changes.
- Unlimited groups and sub-categories
- Descriptions + examples per category
- Synonyms and key phrases
- Drag & drop to reorder
- Versioning and rollback
- Duplicate taxonomies between bots
The AI reads, understands and tags in seconds.
Every incoming message goes through the classifier. One or more categories are applied with a confidence score. Low confidence? Flagged for review automatically.
- Multi-label classification
- Confidence score per label
- Auto-flag for low confidence
- Re-classification when moving categories
- No fine-tuning, no batch queue
- Traceable per conversation
Filter the inbox by category.
In Databot Live, operators filter the queue by group or category. Automatic routes — "all Billing goes to Karla" — fire themselves.
- Persistent filters per agent
- Routing by category
- Different SLA per group
- Targeted notifications
- Saved views per team
- Filtered CSV export
From labels to insights.
Volume by topic, month-over-month trends, top categories with human hand-off, resolution times by group. Your taxonomy becomes your operational map.
- Volume and % per category
- Weekly and monthly trends
- Cross-cuts with channel and sentiment
- Alerts for anomalous spikes
- Export to BI (BigQuery, Redshift)
- Shareable board for stakeholders
The taxonomy travels with the conversation across the entire platform.
From zero to auto-tagging in one week.
Cut through the noise. Discover patterns.
We design the initial taxonomy with you and train the agent to start tagging correctly from day one.