Skip to content
Success Story

How Velotric's AI Agent Handles 58% of Routed Tickets and Cut Agent Research Time From 20 Minutes to Under 5

How Velotric uses Pluno's Deflection AI and AI Copilot inside Zendesk to answer repetitive e-bike questions in seconds and free its 30+ person support team for complex cases.

Cristina Mercado

Cristina Mercado

Reputation and Trust Manager in Customer Support at Velotric

Published

Interview by Alex Abstreiter, CEO & Co-founder at Pluno

Core Challenges

  • Agents spent a significant amount of time answering the same questions again and again, which drove up workload and made it harder to prioritize the complex cases that need a human.
  • Cristina built triggers for common pre-sale questions, like where to book a test ride. The trigger could answer the first question, but the moment a customer asked a follow-up it stopped working, and the ticket had to be reopened and assigned to an agent by hand.
  • The team also tried Zendesk's own AI solution, but the replies did not feel natural. Customers could tell they were talking to a bot.
  • The team was already struggling to keep up with ticket volume. A tool that needed a complex implementation was a non-starter.

Solutions

  • AI Agent for repetitive questions
  • AI Copilot with references
  • Form submission integration
  • Website chat for pre-sale
  • Daily review and coaching
  • Support in Slack

Key Results at a Glance

Before Pluno

Answered by agents

With Pluno

58%

Of Routed Tickets Handled by the AI Agent

Before Pluno

~20 min

With Pluno

< 5 min

Agent Time on a Copilot-Assisted Ticket

Before Pluno

Manual reply

With Pluno

~10 sec

To Answer a Simple Inquiry

Before Pluno

Broke on follow-ups

With Pluno

Full conversation

Pre-Sale Questions

The Challenge

A Seasonal Spike of Repetitive Questions

Velotric's support team of 30+ people answers calls and help desk tickets from direct-to-consumer riders. Biking is seasonal, so from summer through September the team handles around 10,000 tickets a month, and a large share of them ask the same questions.

  • Repetitive inquiries eating agent time: Agents spent a significant amount of time answering the same questions again and again, which drove up workload and made it harder to prioritize the complex cases that need a human.
  • Zendesk triggers broke on the follow-up: Cristina built triggers for common pre-sale questions, like where to book a test ride. The trigger could answer the first question, but the moment a customer asked a follow-up it stopped working, and the ticket had to be reopened and assigned to an agent by hand.
  • AI answers that sounded like AI: The team also tried Zendesk's own AI solution, but the replies did not feel natural. Customers could tell they were talking to a bot.
  • No appetite for a heavy rollout: The team was already struggling to keep up with ticket volume. A tool that needed a complex implementation was a non-starter.

“Our agents are spending a significant amount of time answering questions repeatedly. It increased the workload for our agents, and it is more difficult to prioritize complex cases.”

Cristina Mercado, Reputation and Trust Manager in Customer Support
The Solution

An Instant Setup, Then an AI Agent the Team Calls “Sophia”

Eric Xu, Velotric's co-founder and Head of Customer Experience, discovered Pluno and was impressed after his first meeting with the team. Cristina followed up for a demo and found there was almost nothing to implement: Pluno connected to Zendesk and started working.

Today Velotric runs Deflection AI, Pluno's AI Agent, on routed help desk tickets and, more recently, on website chat focused on pre-sale questions. Agents use the AI Copilot for everything that needs a human.

“There's not much challenge with the implementation. If we're going to activate it, it's going to work. It's like an instant solution. That's what we love with Pluno: it's just a simple integration.”

Cristina Mercado

What Pluno delivered

AI Agent for repetitive questions

The AI Agent answers simple inquiries in about 10 seconds and holds the conversation through follow-up questions, which is exactly where Zendesk triggers used to fail.

AI Copilot with references

Instead of hunting through KB articles and product information, agents get guidance with references next to the ticket, so complex replies go out faster.

Form submission integration

Pluno helped Velotric integrate the form tool customers use to submit requests, and interprets what customers submit.

Website chat for pre-sale

Velotric recently added the AI Agent to chat on its website, focused on questions from riders who are still deciding which e-bike to buy.

Daily review and coaching

Cristina reviews every ticket the AI Agent handles each day, checks the tagging and responses, and coaches it by refining prompts and instructions.

Support in Slack

When Velotric has a question, the Pluno team answers in a shared Slack channel right away, not in a day or two.

“Instead of spending time on repetitive inquiries, our support team can now handle more complex tickets and those that need human intervention. A simple inquiry can be answered in a matter of 10 seconds. That's the best thing.”

Cristina Mercado
The Results

58% of Routed Tickets Handled by AI, and Faster Human Answers

Of Routed Tickets Handled by the AI Agent

Before Pluno
Answered by agents
With Pluno
58%

Agent Time on a Copilot-Assisted Ticket

Before Pluno
~20 min
With Pluno
< 5 min

To Answer a Simple Inquiry

Before Pluno
Manual reply
With Pluno
~10 sec

Pre-Sale Questions

Before Pluno
Broke on follow-ups
With Pluno
Full conversation

Vendor Support Response in Slack

Before Pluno
Days
With Pluno
Instant

Implementation

Before Pluno
Complex rollout
With Pluno
Instant

“Over time, the team became more comfortable, and now they see it as a team member. Before, it was their competitor. If the AI Agent wasn't there, they would be the ones handling those tickets.”

Cristina Mercado, on how her team's view of the AI Agent changed
The Takeaways

“Amazing.”

That is how Cristina sums up Pluno in one word. Customers have even praised Sophia, the AI Agent, by name on Velotric's Facebook page, and her team now spends its time where people matter most.

Next, Velotric wants to widen coverage: when a case needs a quick human decision, like confirming a return, the AI Agent would ask the internal team for a simple yes or no and then continue the conversation with the customer on its own. Cristina's advice to teams with high volume, repetitive questions and a solid knowledge base: pair the Copilot with the AI Agent.

“Pluno helped us turn repetitive questions into scalable AI-assisted support while allowing us to focus our time and expertise where human involvement matters most.”

Cristina Mercado, Reputation and Trust Manager in Customer Support
Interview Transcript

The full interview with Cristina Mercado

Cristina Mercado, Reputation and Trust Manager in Customer Support at Velotric, in conversation with Alex Abstreiter, CEO & Co-founder at Pluno. Lightly edited for readability: filler words and false starts are removed, the wording is as spoken.

Read the transcript13 chapters, 1,751 words

Velotric support role

Cristina Mercado

Velotric is an e-bike company focused on providing reliable and accessible electric bikes directly to consumers and serving a dealer network.

I am working as a Reputation and Trust Manager. I'm responsible for protecting and strengthening the company's reputation when it comes to the customer's experience. I'm monitoring the customer's feedback and reviews, responding to concerns, identifying recurring issues, and coordinating with various stakeholders to make sure we turn the customer's feedback into an improved and positive experience.

Alex Abstreiter

How big is your current customer relationship team that you work on?

Cristina Mercado

The main customer support staff is composed of 30-plus head count. They mainly coordinate with our DTC customers. They are the ones who answer the calls and tickets submitted by the customers to our help desk.

Ticket volume challenges

Alex Abstreiter

And how many tickets or calls do you get on average in a month?

Cristina Mercado

On average, it's a case-by-case basis, as biking is a seasonal activity. During summer up to September, we're getting around 10,000 on average.

Our biggest challenge was the volume of repetitive inquiries. Our agents are spending a significant amount of time answering questions repeatedly. It increased the workload for our agents, and it is more difficult to prioritize complex cases.

I tried activating some of the triggers in our current ticketing tool, which is Zendesk, but it's not targeting much of what we want to address. And it's not giving effective responses to the customer. In the end, it's going to be handled by the agent again.

Zendesk trigger limits

Alex Abstreiter

And what was the specific pain point in the answers that Zendesk in this case gave to customers?

Cristina Mercado

The main limitation with the triggers is when a customer asks the next question. So, let's say for the pre-sale, I created a trigger for, let's say, how can we do a test ride? That means asking for a location. So, if that question is being raised by the customer through our support help desk, I created a trigger with the locations of the dealers for the test ride.

That trigger will answer. But if there is a follow-up question, it can no longer be activated. So, we have to manually open the ticket again and then assign it to the support team.

Discovering Pluno

Alex Abstreiter

And so, you had this pain point and you were looking for solutions. How did you hear about Pluno?

Cristina Mercado

It's Eric who discovered Pluno. Eric Xu is the Head of Customer Experience of Velotric. He's one of the co-founders. Eric is very resourceful, so I think he found it through someone on LinkedIn.

Eric is not easily getting a tool or product, but the first time you met Eric, even he couldn't believe how you did it. He was very amazed with the product. Which is why I reached out to the team to see if I could go ahead and get a demo.

Alex Abstreiter

Can you talk me through the process, what it was like for you after the demo?

Fast setup criteria

Cristina Mercado

Yeah, when I first learned about it from you, we had tried the Zendesk AI solution, but it's not natural. You can see that it's really an AI response. So, when you introduced Pluno, there weren't many challenges with the implementation. It's like if we're going to activate it, it's going to work. It's like an instant solution.

We don't need to do a complex implementation. That gave us the decision to say, let's go for it. Because we want efficiency, we want resolution. We don't want to struggle more.

We're already struggling with how we're going to manage the tickets, and if we're going to be struggling doing the implementation, let's forget it. And that's what we love with Pluno AI. It's just a simple integration.

Alex Abstreiter

Were there any other factors you were evaluating when checking out different tools?

Cristina Mercado

It's easy to use, very simple integration, no other setup needed, and easy-to-reach support.

Alex Abstreiter

Can you maybe tell me an example where you had to reach out to the Pluno team to get support for something?

Cristina Mercado

When we have some inquiry, we can message you through Slack and we're getting an instant response. So, it's not like we have to wait for a day or two. As a customer, that's very important.

AI Agent impact

Alex Abstreiter

Which features do you use most right now?

Cristina Mercado

We are using the AI Agent, and she has helped us a lot. Of those tickets routed to the AI Agent, she covers at least 58%, which lessens the workload of our support team.

Alex Abstreiter

How did you feel this lighter workload on the team?

Cristina Mercado

Because instead of spending time on repetitive inquiries, those simple inquiries, our support team can now handle the more complex tickets and those that need human intervention. Simple inquiries can be answered in a matter of 10 seconds. So, that's the best thing.

Copilot speeds replies

Alex Abstreiter

And then your team is also using the Copilot to help them with these more complex inquiries?

Cristina Mercado

Yes, and it increased their efficiency. Before, they had to search for the KB articles and search for the product information. Right now, with the AI Copilot guidance providing the references, it helps the team to be more efficient and fast, providing fast responses to our customers.

Alex Abstreiter

Have you checked the time to resolution or time to first response, how it changed over time?

Cristina Mercado

For those agents that have been handling tickets using the Copilot, before, they would spend around 20 minutes, because they had to search for the KB articles. Now, they can do it in less than 5 minutes.

Integrations and tools

Alex Abstreiter

How does Pluno integrate with your existing tools?

Cristina Mercado

Yes, it's with our existing tool, Zendesk. It's easily integrated. It's working smoothly.

Pluno also helped us with the integration of one of our tools. It's called Hark. It's a form of submission as well, and it can interpret what submissions are made by the customer.

Alex Abstreiter

And how does your team feel working with Pluno as a tool?

Team adoption journey

Alex Abstreiter

When you basically started trialing Pluno and gave it to your team, what was the initial reaction, and what is the sentiment right now?

Cristina Mercado

Yeah, at the very first, since they know it's an AI, there was some natural caution. At first, they were hesitant to use it because they were thinking it would replace the human job. Over time, the team became more comfortable, and they now see it as a team member. Before, it was their competitor.

They learned that if the AI Agent wasn't there, they'd be the ones who would handle those tickets. Imagine 50% capacity is being handled by the AI Agent. They no longer need to handle those, so it lessened their tasks.

Alex Abstreiter

It's also important from the management perspective to make this clear for the team, that this is not about replacing them, but more about working together with the tool.

Cristina Mercado

Yeah, we explained everything: this is an AI age, an AI era. We need to adapt. All we need is feedback, so we can help the implementation go more smoothly and be able to address the things needed by the AI Copilot for us to work more efficiently.

Customer feedback wins

Alex Abstreiter

Now, switching from the team more to the customer's side of things, what was the feedback that you received from customers about the support experience that is more powered by AI now?

Cristina Mercado

We're getting a lot of good feedback on our AI Agent. You can see it even if you visit our Facebook page. Our agent is being mentioned as an efficient agent.

From the tickets that she handled, the AI recognition, or feedback that it was an AI, is less than 20. That's how good it is. Yeah, it's a very human tone. It's like I just give the right prompt and then she follows naturally.

Alex Abstreiter

Awesome.

Coaching the AI daily

Alex Abstreiter

How do you currently manage this loop of reviewing conversations that the AI Agent is having and then adapting prompts or workflows on the Pluno side?

Cristina Mercado

Yeah, for me it's a daily review, even the flagging of the responses. I just do a ticket scrubbing of all the tickets she handled. The coaching for the AI Agent is giving the right prompt and providing the right instruction.

Alex Abstreiter

So, we have an overview of how you use Pluno with the AI Agent and also the Copilot. We recently also started the chat directly on the website, more focused on the pre-sale aspect, right?

Cristina Mercado

Yes.

Alex Abstreiter

How would you describe your experience working with the Pluno team?

Cristina Mercado

Pluno helped us turn the repetitive questions into scalable AI-assisted support, while allowing us to focus our time and expertise where human involvement matters most.

Future goals and escalation

Alex Abstreiter

And looking ahead, what are your future goals with Pluno, Cristina?

Cristina Mercado

We want to have bigger coverage for Pluno, as much as possible, but where it can be answered by the internal team. If we can integrate it into our messaging so the AI Agent can continue those types of inquiries, that would be awesome.

Alex Abstreiter

Okay. So, you basically mean a user reaches out, and the AI Agent knows, 'Hey, this is a case where I need some kind of human decision input, for example, to confirm whether they get a return or whatever it is.' And then Pluno would escalate this to your internal team, wait for their feedback, for example, just a yes or...

Cristina Mercado

Yes.

Alex Abstreiter

And continue communicating with the customer in the end.

Cristina Mercado

That's true, because you don't need to provide responses. It's just confirmation. It's just yes or no. So, the agent doesn't need to handle those types of tickets. It's not going to add to their tasks.

Recommendation and wrap

Alex Abstreiter

Now, with all this experience you had with Pluno, what would you say to another company that is considering Pluno right now?

Cristina Mercado

I would recommend it. As we are approaching the AI era, Pluno will be the best support AI as a partnership, because it starts with a very easy integration and a very supportive Pluno support team that's been there, very responsive.

Alex Abstreiter

From your perspective, what type of teams benefit most from Pluno?

Cristina Mercado

Companies with high volume and those repetitive inquiries could benefit. Especially if they have an established process and a strong knowledge base, it will be more efficient to have Pluno. If you put in the Copilot guidance and the AI Agent, it will be more efficient for the company.

Alex Abstreiter

If you had to describe Pluno to a friend in one sentence only, how would you describe it?

Cristina Mercado

Amazing.

Alex Abstreiter

Okay. Thanks so much, Cristina, for this short conversation. I'm looking forward to continuing this partnership with Velotric.

Cristina Mercado

That's honest feedback, by the way.

Alex Abstreiter

Thanks.

Stop settling for AI that needs constant babysitting

See how Pluno resolves 65% of tickets on average across customers (as of November 2026), including complex cases, while supercharging agent productivity.