All use cases
SupportAnalysis

See request topics and where the load is growing

Shift planning runs on a hunch that Mondays are hard. The agent shows which topics and hours carry the load, and where it grows week over week.

Medium · 20 min · once a month

What you get

Requests by topic · 8 weeks · week-over-week change
Delivery status   412 → 517   +26%   peaks Mon 10-13
Returns           188 → 201    +7%   flat across the week
Sizing             96 →  94      —   no change
Payment failed      —  →  61   new topic from week 14
   → needs a person: the week-3 spike came from one mailout

A sample on made-up data — your numbers will be your own.

Who it fits

  • It is time to rebuild the shift schedule, and the decision rests on a hunch that Mondays are hard.
  • The team is asking for one more operator, and you need to show which topic and hours create the growth.

When it won't work

Less than six to eight weeks of history — on three points a trend is indistinguishable from a spike.

How the agent does it

1

Set period and shifts

Take at least eight weeks and name your shift boundaries and time zone. If a topic list exists, hand it over — otherwise next month will not line up.

2

Run the count

The agent counts one request per conversation rather than per message, and separates one-off spikes from steady growth.

3

Match to the roster

These numbers move real people around the schedule. The cost of an error is not a wrong report but an empty or drowning shift.

What you set

Connect Telegram, Avito and email. By hand: the period to count, your own list of topics or consent to the agent's, and your shift hours.

What you'll need

Telegram
Avito
Почта

Starter prompt

Copy the prompt or open it straight in a chat with the agent.

Prompt for the agent

Sort the tickets from Telegram, Avito and email for [period] by topic and count the load by week. Take the classifier from [the attached list of topics] or build your own and show it to me before the analysis. One ticket — one topic; a conversation stretching over several days counts once, by the date of the first message. For each topic: tickets per week, change against the previous week, distribution by day of week and by hour [time zone]. A spike caused by a promotion, an outage or holidays goes on a separate line and is not mixed into the trend. Output a "topic × week" table, a "day of week × hour" table and the three fastest growing topics.

Open in chat
Check the result

Check the load calculation. List: — topics that got fewer than [20] tickets over the period: their weekly dynamics mean nothing; — tickets that ended up in "other", and what they have in common; — conversations counted twice because the customer replied on different days; — weeks with holidays or an outage that must not be put into the trend alongside ordinary ones; — hours where the peak is explained by a single mailout rather than a steady flow; — what the data already shows for the shift schedule and what needs another [N] weeks of measurement.

A second prompt — the agent uses it to review its own work and show what's left for you.

Related use cases