Hacker News Reader

Comment by inetknght | Hacker News Reader

Many companies use this style of "support" in place of employing real people and don't provide a method to escalate to a real human. What guard do I have, as a consumer, against trash like that? What makes yours unique?

When I call for support I expect to receive support. I have never ever had an AI or robot support assistant that actually provides the help that I need when I call.

How do you address the possibility of LLM generating absolutely false information?

Do you actively tell the user where they can contact a real human to provide feedback about the terrible customer support response that your product will undoubtedly provide? Or is this just another dumpster fire of "it sounds like you're having trouble with our product. here's our FAQ..."?

Does your product integrate with internal company APIs? How do you deal with the risk of customers abusing that?

View on HN

We have a number of safeguards:

1) Our customers can decide which categories of end-user questions should be auto-escalated

2) We ensure that if a question cannot be answered based directly on knowledgebase content, we either escalate or try to ask followups to drill deeper

3) We offer an API which our customers can directly integrate with their own systems. For protection, we either limit the set of allowed actions, or if the customer prefers, they can handle the safeguarding code internally with advice from us.

Lastly I would say the usage of LLMs is young in the support space and things will only improve from here!

Most of this AI stuff is sorely lacking a "feedback loop" If you aren't collecting data on angry/ unhelped customers... you are just flushing your business down the drain

Also maybe it is just me... but I would never call customer service unless i have already exhausted all self help options, meaning i need a real human who can poke into the database or whatever to fix a real edge case problem, who did the calculations to decide these power users (best customers?) can all just die?

How much do you plan to pay for the product? If you're paying a few dollars per month for a consumer product, be aware that having a human answer a single inbound call or email from you can often wipe out the entire year's revenue (not just profit) the company is getting from you.

Either we all need to be willing to pay more in exchange for "premier" support being available or we need to be ok with companies trying to cut down the support load.

> Either we all need to be willing to pay more in exchange for "premier" support being available or we need to be ok with companies trying to cut down the support load.

Third option: if you can't afford to provide human support then you don't deserve to be a business.

Too expensive to provide support? Then raise your prices. If your customers leave then your product isn't viable.

Ok but you’re still making the same point. You can’t have something cheap with a white glove CS experience for every caller. Either you have to have a screen to filter out the complete idiots who will waste their time all day because their mouse isn’t plugged in, like AI, or some minimum wage tier 1, etc. Or it has to be cost more to pay for that cs.

> Ok but you’re still making the same point. You can’t have something cheap with a white glove CS experience for every caller.

That is my point.

The "cheap with a white glove CS experience" is a completely shit experience. If you can't afford to pay for real customer support then either raise your prices or go bankrupt.

Building a business without providing customer support should not ever be a valid business model.

Reply on news.ycombinator.com