Know which ticket needs you — before the customer chases
One screen that tells you what's at risk, who's unhappy, how fast you're replying, and how much the AI handled. Then a copilot you can simply ask.
Today's briefing
Quieter week — 34 tickets in, 31 resolved. The AI handled 19 of them on its own, mostly billing and password questions. Two tickets have been waiting over a day: #4802 (checkout errors, urgent) and #4795, where the customer has now replied three times. Doc helpfulness dipped on "Setting up payouts" — six people marked it unhelpful.
Customer sentiment
- Positive
- 68%
- Neutral
- 24%
- Negative
- 8%
Open tickets
Waiting on us
Median first reply
Resolved · 7 days
Docs published
Doc helpfulness
AI deflection
AI replies sent
TICKETS
Checkout errors after update
Payout still not received
License activation failing
Can I import from SupportCandy?
Two sites, one license?
Chatbot not showing on mobile
How do I brand the chat widget?
Refund policy question
Feature request: email intake
AI PERFORMANCE
- 19
- replies sent
- 56%
- deflection
- 11
- stayed quiet
- 284
- questions
- 91%
- answered
- 26
- handed to a human
Illustrative data. Your dashboard fills in from your own tickets, docs and AI activity.
The 360° view
Six questions it answers before you ask
Most help desks show you a list of tickets and leave the thinking to you. This one starts with the judgement calls.
Which ticket will blow up if I ignore it?
The needs-attention queue ranks by risk — how long someone has waited, how many times they have followed up, and how the thread reads.
Which customer is unhappy right now?
Replies are scored as they arrive, so negative threads surface instead of hiding at the bottom of a list sorted by date.
Are we actually replying faster?
Median first reply and resolution counts, compared against the previous period, so you can see the trend rather than today’s mood.
Is the AI earning its keep?
Deflection rate, replies sent, and how often it chose to stay quiet — the numbers you need to decide whether to loosen the threshold.
Which help article is failing people?
Article votes are the reader signal NiroHelp records, so the not-helpful count tells you which page to rewrite first.
What happened while I was away?
An activity feed across tickets, docs and AI, filterable, so a Monday morning catch-up takes a minute.
Ask NiroHelp anything about your support
The copilot sits in your dashboard toolbar and reads your own tickets, replies and help articles. Ask it in plain language — no filters to build, no report to configure.
It answers from your data and cites what it looked at. When it can't tell — because the history isn't there yet — it says so instead of inventing a number.
Based on 12 open tickets and 40 replies in the last 30 days
Based on doc votes and ticket references
Try asking
- Who has been waiting longest?
- How did we do vs last month?
- Which agent is overloaded?
- Should I raise the AI threshold?
What's on the screen
Every panel, and the decision it's there to support.
Today's briefing
ProA written summary of the day: what came in, what the AI handled, what is stuck, and which article slipped.
Answers: where do I start today?
KPI tiles
FreeOpen tickets, waiting on us, median first reply, resolved, docs published, doc helpfulness, and AI activity — each against the previous period.
Answers: are we getting better or worse?
Needs attention
FreeTickets ranked by risk rather than date, combining wait time, follow-up count, urgency and reply tone.
Answers: what will hurt if I ignore it?
Customer sentiment
ProA positive / neutral / negative split across recent replies, scored locally as they arrive.
Answers: who is unhappy right now?
AI performance
ProReplies sent, deflection rate, how often the AI stayed quiet, and chatbot volume — with a link straight to the threshold setting.
Answers: should I trust it with more?
Recent activity
FreeA single feed of ticket, doc and AI events, filterable by type.
Answers: what happened while I was away?
Common questions
No. The dashboard — KPI tiles, the three ticket queues, activity feed and doc metrics — comes with the free plan. The copilot and the AI performance panels need Pro, because they depend on the AI features.
It weighs how long the customer has been waiting, whether anyone has replied yet, the urgency on the ticket and how the recent replies read. Tickets rise to the top when several of those line up badly.
Replies are scored locally as they arrive and rolled into a positive / neutral / negative split, so you can see the mood of the queue rather than reading every thread. It is a signal to act on, not a score shown to your customers.
From your own data — tickets, replies, agents, help articles and AI activity in the range you have selected. It does not reach outside your site for business answers, and it cites what it looked at.
Panels that need history say so plainly instead of showing a fake zero-line. Metrics like median first reply and AI deflection fill in as your team works, usually within the first week.
Yes — the range selector at the top drives every panel on the page, so switching to the last 30 days recalculates the tiles, the queues and the AI performance numbers together.
See your support at a glance
The dashboard comes with the free plan. The copilot arrives with Pro.