.png){kind=small}
Why it matters
You can’t improve what you can’t see. The Insights Dashboard translates thousands of searches and conversations into clear metrics and action lists—so marketing, product, and support teams can move together.
From raw chat logs to revenue‑level signals.
What it is
A real‑time analytics workspace with filters, drilldowns, and exports. Track answer quality, sentiment, demand, geography, timelines, and product mentions—then jump straight to the underlying threads.
Highlights:
-
Auto‑flag conversations that need human review (policy risk, incomplete answers, unhappy feedback)
-
Visualise answer completeness and agent mix at a glance
-
See campaign impact, sentiment, geo hotspots, and product trends
-
Export charts and tables for decks and BI
What you’ll see
Top tiles
-
Total Outcomes – count of solved actions/answers
-
Fully/Partially/Unanswered – completeness breakdown
Charts
-
Answer Completeness – donut chart of fully/partially/not answered
-
Expert Agents Mix – which task‑specialist agents handled demand (e.g., Tech Support, Sales, Dealer Finder)
-
User Feedback – positive vs negative ratings
-
Sentiment Distribution – non‑neutral sentiment over time
-
Billable Interactions – billable vs non‑billable counts
Tables
-
Products Mentioned – ranked list of SKUs/models by conversation count
-
Threads – filterable list with tags (e.g., Tech Support, Compatibility) and quick open for full context
Maps & Timelines
-
Geography – regional hotspots so you can aim service, inventory, and campaigns
-
Timelines – trend lines for total threads and completeness status month‑over‑month
Smart triage with human‑review flags
The dashboard automatically flags conversations that are likely to benefit from human attention—such as unclear answers, negative or conflicted sentiment, or safety/brand policy triggers. Review queues are filterable and exportable, so owners can fix content, update FAQs, or follow up with customers.
Filters that match your business
Slice insights by:
-
Date range
-
Products / SKUs
-
Question Type (e.g., Tech Support, Sales, Compatibility)
-
Answer Completeness
-
Billable status
-
User Rating and Sentiment
-
Needs Review (human‑flag queue)
Every filter updates charts, maps, timelines, and tables instantly.
From insight to action
-
Identify knowledge base gaps and publish targeted FAQs
-
Track campaign impact from in‑conversation mentions
-
Spot geographic demand for better store placement and staffing
-
Quantify product mentions to inform inventory and roadmap
-
Monitor product recall awareness and messaging effectiveness
Governance & accuracy
-
Source‑level permissions and audit trails
-
Conversation transcripts with citations back to the exact sections used
-
Privacy controls and retention windows (AU/NZ data options available)
Implementation
-
Setup: Included with Airgentic Search and Agentic AI deployments
-
Time to value: Prebuilt dashboards live on day one; customise via filters and exports
-
Exports: One‑click PNG/CSV for decks and BI tools
FAQs
Is this only for support?
No. Product and marketing teams use it for demand signals, content gaps, and campaign impact.
Can I see the raw conversations?
Yes—open any metric to jump to the exact threads with timestamps and tags.
How does auto‑flagging work?
We combine answer quality checks, rating/sentiment signals, and policy heuristics to queue items for review.
Do we need a BI team to use this?
No. It’s point‑and‑click with export options for deeper analysis.
Get Started
Contact us for a live demo.