Ali Merchant is the Founder and CEO of Orvera AI, a San Francisco company that helps brands run customer experience with AI agents. Orvera’s AI agents resolve customer conversations across voice, chat, text and email, and they share one platform with real-time assist for human agents, quality scoring on every conversation and voice-of-customer analytics. Orvera builds, deploys and runs the agents for its customers, and a full deployment averages three to six weeks.
Ali has spent 18 years running contact-center and health-information operations. In 2008 he founded AM Infoweb, a health information and contact center operations company, and he still leads it today. That time on the operations floor shapes how Orvera works: the company takes responsibility for the agents it puts live and judges them on whether the customer’s problem got solved. Orvera works with brands in e-commerce, travel and hospitality, and insurance across North America.
Company: Orvera AI
We are thrilled to have you join us today, welcome to ValiantCEO Magazine’s exclusive interview! Let’s start off with a little introduction. Tell our readers a bit about yourself and your company.
Ali Merchant: Thanks for having me. I’m Ali Merchant, Founder and CEO of Orvera AI. We help brands run customer experience with AI agents. Our AI agents talk with customers on the phone and over chat, text and email, and they’re built to finish the job: track the order, change the booking, open the claim.
Everything sits on one platform. The same system that runs the AI agents also coaches human agents live on their calls, scores every conversation for quality and pulls out what customers are telling the business. We build it, deploy it and run it for our customers.
I came to this from the operations side. In 2008 I founded AM Infoweb, a health information and contact center operations company that has grown to more than 2,200 employees, and I still lead it. Eighteen years of running those floors showed me how much work a customer team quietly does to keep its tools running. We started Orvera to take that work off their plate.
What was the moment you realized that complexity was holding your company back, and how did that shift your approach to product, operations, or strategy?
Ali Merchant: It came when we were building a medical records review and summary platform. We came from that industry, and we got the platform to 86% accuracy, which was the best number we’d seen anywhere. Other platforms were sitting at 70 to 80%. We never launched it. We’d done that work by hand for years, so we knew 86% doesn’t cut it on medical records. It has to be at human level, 98 to 99%, or it isn’t usable.
What I took from that is where the real complexity sits: with the customer. An AI that gets it right most of the time quietly creates a second queue of cleanup work, and someone on the customer’s team ends up owning it. So we started asking one question about every product decision. Who does the work? At Orvera, we do. We build the agents, connect them to the customer’s systems, put them live and keep running them. The customer gets the outcome and we carry the project.
Can you share one specific example of something you simplified (a product feature, process, pricing model, or customer experience) that delivered outsized results?
Ali Merchant: The biggest one was the deployment itself. A typical AI project in a contact center runs six to nine months. A vendor sells the software, an integrator wires it up, and the customer’s own team carries the testing and tuning. That’s three parties and a lot of handoffs.
We pulled all of that into one team. We build, integrate and operate the agents ourselves, and across our customers a full deployment averages three to six weeks, end to end.
The second thing we simplified was putting everything on one platform. The AI agents share one memory of the customer across channels, so someone who texts and then calls doesn’t have to start over. When a conversation moves to a person, that person gets the context and a summary, and Agent Assist puts the next answer in front of them while the customer is still talking. Across our customers, first-contact resolution runs at roughly 80%, and handle time on the human side falls 15% within the first 90 days.
How do you systematically identify and eliminate complexity across the organization without oversimplifying in ways that hurt quality or value?
Ali Merchant: We look at it from the customer’s side of the conversation. Most contact centers only ever review a small sample of their calls, so the complexity hides in the rest. We score every conversation on every channel, which shows us exactly where customers repeat themselves, wait on a transfer or call back about the same problem. Each of those is a place where things got harder than they needed to be.
The guard against oversimplifying is that we measure whether the problem got solved. Ending a conversation quickly is easy. If an agent closes a call fast and the customer calls back the next day, nothing got simpler. The work moved to tomorrow. So every change gets judged on resolution.
We also train contextualization models on each customer’s de-identified data, so the AI agents understand that business’s own products, policies and language. And when something is genuinely complex, like an exception that needs judgment, the agent hands it to a person with the full context.
What’s the biggest resistance you’ve faced when pushing for radical simplicity, and how did you overcome it?
Ali Merchant: Honestly, the resistance comes from people who are good at their jobs. Contact-center leaders have been sold a lot of tools, and plenty of them have lived through an AI project that took most of a year and still needed their team to babysit it. So when we tell them we’ll run it for them and they’ll be live in weeks, they’re skeptical. They should be.
What works is proof. We start with one use case they care about, like order status for a retailer or a first notice of loss for an insurer, put it live and let them watch it on their own customers. Once a team sees an agent handle real conversations and pass the hard ones to their people cleanly, the conversation changes. I’d rather earn that trust one use case at a time than win an argument in a meeting.
For other CEOs who struggle with feature bloat, process overload, or confusing customer journeys, what’s the most important principle or first step you would recommend they take?
Ali Merchant: Start by counting who has to do what. Pick your most common customer request and write down every step, every system and every person it touches before it’s solved. Most CEOs are surprised by the number. Then ask which of those steps the customer actually cares about. Usually it’s one, getting the problem fixed.
For anything you build or buy, ask the question we ask ourselves: once this is live, who’s doing the work? If the answer is your team, with a new dashboard, a new login and a new set of exceptions to manage, you’ve added complexity, even if the demo looked simple. The simplest thing you can give a customer, or your own team, is fewer things to manage.
