Reinventing Customer Service

Check out this headline,  

Klarna’s AI Assistant Is Doing the Job Of 700 Workers, Company Says 

and conspicuously followed by,  

'But Job Replacement isn't the Biggest Issue.' 

and the Pope is Punjabi. 

If you are a brand, let’s say a Telecom or a Bank, and you talk about quality and productivity. What you rightfully mean is how to dispose of the customer soonest? 

That is a good aim to have.  

One. Customer service is a horrible job. Customer service workers have the highest turnover rates ~ 30% on average. Harvard Business Review.  

Two. A satisfied customer is a happy customer. 

Happy customers make for good business.  

And happy employees more so. 

But what makes the shareholders happy is happy customers without the happy employees.  

Welcome to the world of Large Language Models and Artificial Intelligence 

But before that.  

Resetting the Score 

Sometimes, an entire industry gets reset to zero, and all the entrenched advantages and parameters go away. The iPhone had that effect, and so did HMS Dreadnought.   

I would urge you to read this article here, 

If I were to summarize, the point here is that there is a huge reset that has happened, and if one is resourceful and a bit of a techno brigand, you can leverage huge technological advances to build software tools and products at a fraction of the cost.  

And if I dare say at one-tenth of the required manpower.  

HMS Dreadnought was the first modern battleship.  

iPhone was the first time; personal computing was first set to zero.  

OpenAI, and OpenAI inspired large language models would make the biggest dent in customer service and support.   

It will make more so in education technology (EdTech) but that is for another time.  

Still,  

If we dig deeper and read this Wired article.

There are gimmicks galore but some serious projects; them or versions of them are bound to succeed. Ease of use, convenience, and cheap. This combination usually works. Not very much unlike Craigslist taking newspaper's classifieds business away and deep within the rock, they buried it. In a tomb so dark, it would never come to light, and it never did. 

AI problems are Data Problems. 

Customer Service and Communication is Contextual Conversation. 

That statement is mouthful, and you need an intelligent LLM to parse it for the reader.  

Jokes aside, LLMs interact with your data - it can be billing, complaints, scheduling, and searching; although LLM search is more about answers, again later on that, service issues, rates negotiations, customer acquisition, translations, a bit of HR, approvals, also disapprovals and the list is endless.  

For technology companies (is there any other kind?) these are overheads and if you are not building and if you are not selling; why do you exist?