About the client
Our client is a large technology company that receives about 1,500 spam and abuse cases every day. Each one needed the same routine: read the complaint, check the IP address, suspend the account and send a closing email. That took 3 to 4 minutes a case, mostly lookups and clicks. Ten agents cleared roughly 300 to 400 cases a day, so work spilled late into the night.
What we did
- 1
On arrival, an AI model flags spam, reads sentiment, summarizes the case and writes its verdict to the record.
- 2
A scheduled RPA bot signs in to the network security console and blocks any flagged IP address not yet blocked.
- 3
The CRM routes each account suspension by ownership. For one user directory, the suspension goes through an API. The other directory has no API, so an admin RPA bot does it.
- 4
A scheduled job closes the case, queues it for audit and sends the matching email template. This step uses native CRM automation, with no AI or RPA needed.
- 5
Every status change is written to the case, feeding automated dashboards.
Facing similar challenges in your CRM?
The results
Figures are as measured for this implementation and are not a guarantee for other teams.
*Delivered as Vegacompass.
