Arsen Misakyan on Why AI Won’t Make the Hard Calls

Share this blog via -

Summarize this post with:

Humans of Support is our interview series featuring the people building and running customer support teams. Each episode brings a different perspective from the industry. 

Every conversation runs in two parts. First, questions built around that guest’s specific expertise and experience. Then, a set of rapid-fire questions we ask every guest to get a good picture of the industry landscape. 

For episode two, we’re talking to Arsen MisakyanFounder and CEO of LAXcar, a Los Angeles-based transportation company running airport transfers, sedans, SUVs, Sprinters, coach buses, and private tours. Before LAXcar, Arsen spent 11 years in customer support at HSBC.

In this conversation, we asked Arsen what changes when you go from a scripted, compliance-driven environment like banking to a business where nothing can be scripted, what happens behind the scenes the moment a flight gets delayed, and why he thinks the idea of AI replacing human judgment in logistics is overhyped. 

You spent 11 years in customer support at HSBC before moving into transportation. What's the biggest difference between banking support and support at LAXcar, and how has your approach evolved from one to the other?

Customer support in banks is extremely structured: all communications are scripted, go through compliance filters, and fall into the escalation tree. At LAXcarit’s changing literally every minute: delays, drivers stuck in traffic, customers waiting at the airport terminal, nothing can be scripted. 

"I learned that the key to good customer service is solving issues here and now."

Arsen MisakyanFounder and CEO of LAXcar

LAXcar today does many things beyond airport pickups: sedans, SUVs, Sprinters, coach buses, private tours. Which is the most challenging to support and why?

The most difficult services to support are definitely coach buses and Sprinters for large groups. Organizing 20 to 40 customers with varying arrival times, luggage amounts, itinerary changes, and other aspects, all fitting into one vehicle with one driver, is impossible without mistakes. 

What happens when someone makes changes to a booking, something like "my flight's delayed and I need my pickup pushed"?

This is taken care of in real time. We immediately contact our dispatcher; the driver is either redirected or stopped, and thereafter the customer gets confirmation before hanging up. 

"This is why flight monitoring is one of the processes in the flow."

What's your promise to a customer, for example, on-time pickup, and what happens internally when you're about to miss it?

We assure the customer that their ride will be there before they arrive. In case of missing the guarantee, the company immediately escalates by contacting the driver and finding a replacement if needed, as well as informing the client. 

What's the most overhyped claim you hear about AI in support or logistics right now?

The claim that AI will replace humans in making critical logistics decisions. 

"It will not. AI will provide additional information, but making decisions when three problems occur simultaneously will always be a human's task."

From 2006 to 2016 to 2026, how have you seen support change?

In 2006, support was phone-based with a large paper trail. In 2016, apps and real-time tracking arrived, so customer expectations skyrocketed. In 2026, support is assisted by AI, but expectations about personal communication become even higher: when a client calls, they expect something special. 

If you could change one thing about how companies think about customer support, what would it be?

"Support is not a cost. Support is your link to the reality of customer experience."

People who start working in it right away progress quickly and stick around for a long time. 

AI: boon or bane for a business built on the human touch, live 24/7 reps?

Boon, if used wisely. AI does the boring stuff: confirmation, tracking, simple updates, leaving the most interesting tasks for our live agents, who take the decision-making process into their own hands.

People who start working in it right away progress quickly and stick around for a long time. 

If you had to describe great support, the kind LAXcar tries to deliver, in one sentence, what would you say?

"Great support is when customers do not have to worry about the service they paid you to perform." 

What matters more: acknowledging a customer request fast, or actually resolving it fast?

Resolving the problem. Acknowledging the request without solving it only creates noise. 

What metric do you watch most closely, and what does a bad week look like in that number?

The on-time pickup rate. A bad week means any performance lower than 97%. Any decrease in this number is followed by an analysis within 24 hours. 

Do you think a human on the phone will still matter in five years, no matter how good AI gets?

Yes. Not all cases require it, but it becomes necessary sometimes: a late-night pick-up, a worried client, a wedding-day problem. 

"People need a comforting voice to talk to."

How do you handle support queries differently when it comes to a customer vs. the chauffeur?

Customers need a more comforting and solution-oriented approach. Chauffeurs require only the correct information without unnecessary talking. 

From 2006 to 2016 to 2026, how have you seen support change?

In 2006, support was phone-based with a large paper trail. In 2016, apps and real-time tracking arrived, so customer expectations skyrocketed. In 2026, support is assisted by AI, but expectations about personal communication become even higher: when a client calls, they expect something special. 

That's episode two of the Humans of Support Series. If you're building or running an IT support team and have a story worth telling, we'd love to hear from you.

Table of Contents

Trusted by 7,000+ customers

Need assistance with Desk365? Get started with our articles in the Help Center!

Accordion Content

                   
Descriptive Image Text

Quick Overview of Desk365's Agent Bot and Support Bot for Microsoft Teams

Watch video

Getting Started with Desk365: Your Modern Helpdesk Ticketing System

Watch Video

Descriptive Text Here