What's everyone accomplishing (or working to do) with AI?
I'm still dubious on AI as a whole (warming up, but pissed about RAM prices LOL), but have used it to good success with a few tasks in our org.
OPERATIONS:
- Monitoring inbound ticket queue, flag and set priority on tickets based on analysis of the issue. For urgent/critical issues, it messages the "everybody" channel in our slack with a big bold alert. Helping to make sure we jump on important issues quickly.
- Analyzing ticket trends, proactively alerting management to abnormalities, response time issues (this weeks response times significantly worse than last week), identify most efficient agents, etc., even identifying blockers on tickets open longer than usual.
- Additional issue research for tickets when techs are "stuck", they can ask the AI to research a specific ticket. It reads the ticket details and notes, puts together a troubleshooting guide, pulls relevant information from our documentation, and pulls relevant info from the internet (message boards, documentation, etc). This has been far more effective at ID'ing some tough issues that Google or ChatGPT have been since it can get full context from the ticket. Of course, it's dependent on the tech taking good notes of their troubleshooting steps so results vary.
SALES:
AI note taking app used in all sales meetings (either Zoom or in person, have been trialing a few different AI note taking tools). Can provide later analysis of things like how their current security measures align with specific regulations and standards, flag action items, pain points, etc.
Take those AI notes and dump them into another AI tool we use (same one that monitors our ticket queue) and it pumps out fully built, nicely formatted proposals based on the pain points, security & operational gaps, and solutions discussed in the meeting. It automatically goes out to the potential client's website and grabs their logo, puts that on the cover page, and spits out a ready to edit Word doc. Takes our proposal creation from hours down to a few minutes to generate + maybe 15 minutes of editing the word doc.
AI monitors incoming emails to a specific mailbox containing order status updates, grabs the PO section of those notifications, matches it to the PO field in the associated order ticket, updates a note in the ticket with the relevant information (what items have shipped, what have not). Working on getting it to integrate with the Ingram API next for better integration as the email integration has been just ok.
We're also talking about ways to integrate AI into project management, haven't started that initiative yet. We use Zoho projects currently and have tied AI into it to test, it can grab relevant status and post it in a ticket on a schedule to keep a ticket and a project "in sync". Haven't implemented.
So far we're impressed with what we can do with AI, but also concerned with how quickly a single prompt can gather a bunch of information that could be dangerous in the wrong hands. Trying to balance convenience with safety. Definitely NOT giving it access to any passwords!
Curious what others in ownership/mgmt are doing to use AI to help augment staff and take time consuming tasks off their folks, reduce missed communications, reduce redundant work, etc. Of course, we are concerned about potential security risks, making sure we only engage with AI products that are SOC2 compliant, don't train on our data, limit access scope, etc.