In this episode of Born In Silicon Valley, Orvera AI Founder and CEO Ali Merchant joins host Jake Aaron Villarreal to discuss how AI is changing enterprise customer experience and contact center operations. Drawing on nearly two decades of experience building and managing contact centers, Ali explains why the technology is ready for enterprise use, even though adoption remains limited.
The conversation covers an omnichannel benefits administration deployment that handled 84% of calls, emails, and texts, the importance of measuring conversation resolution, and why industry context will separate lasting AI platforms from short-lived products. Ali also discusses human handoffs, hiring, bootstrapping, Voice of Customer intelligence, automated quality management, and the shift from contact centers as cost centers to command centers.
Ali Merchant:
One of our customers said, “Now that we have AI agents, can’t we make 400 outbound calls at once?”
We said, “Yes, technically you can. However, you may end up flooding payer lines and damaging those relationships.”
The objective may be to get paid faster, but excessive automation could result in delayed payments, damaged relationships, or other challenges.
Just because the technology allows you to do something does not always mean you should do it.
Jake Aaron Villarreal:
I’m Jake Aaron Villarreal, born and raised in Silicon Valley. This show takes you behind the scenes to understand what it is like to be a startup founder, the journeys founders are on, the problems they face, and the products they build to improve our lives.
I’m excited to welcome Ali Merchant, Founder and CEO of Orvera AI.
Ali, welcome to the show.
Ali Merchant:
Thank you so much. Thank you for having me.
Jake:
We have been trying to arrange this conversation for a while, so I’m glad we are finally here.
Where are you joining us from today?
Ali:
I’m joining you from our Pune office in India.
Jake:
I love Pune. We also have an office there. There is a great deal of technology and innovation happening in the city.
For anyone unfamiliar with Ali and Orvera AI, Orvera is a conversational AI platform specializing in voice AI for enterprise contact centers.
Ali believed early that contact centers would eventually buy complete AI platforms rather than isolated point solutions. That understanding came from spending 18 years building AM Infoweb, a company with approximately 2,000 employees that helps organizations establish and scale contact center operations.
Before we discuss Orvera AI, tell us more about your background. How did you enter technology, and what shaped your journey?
Ali:
I grew up in a family business. Family businesses are often traditional and built around established ways of working, but I wanted to create something new.
After completing my education, I started AM Infoweb. I ran the business for 18 years and met several great people along the way who became co-leaders within the organization.
When ChatGPT arrived, it created a major shift for us.
We had always known that AI was developing. We had heard about it, experimented with it, and followed the technology. However, the moment we began using ChatGPT, we understood that the world was about to change in a way we had not seen before.
We knew we needed to act quickly and seriously to make the most of the opportunity.
We created an R&D team and brought together some of the strongest technology professionals from AM Infoweb. We also included experienced contact center leaders.
It was a natural decision. If you are building a platform for contact centers and contact center leaders, those professionals need to be part of the team developing it.
That is how Orvera AI began.
Jake:
People often think of a contact center as a place they call when they need help with a credit card, a bank account, or a technical problem.
You have worked in this space for many years. What problems did you see that made you believe AI could make a meaningful difference?
Ali:
Through AM Infoweb, we support enterprises with in-house contact centers as well as outsourcing companies that run contact center operations for their clients.
We help them establish, manage, scale, and transform those operations.
That gave us exposure to multiple industries and a wide variety of use cases.
Over the years, every new technology promised major improvements. In my personal experience, many of those technologies took time before they delivered on that promise.
AI was different.
Right from the beginning, it showed significant potential.
Contact center leaders have lived a difficult operational life. Many contact centers operate 24 hours a day, seven days a week, 365 days a year.
As a leader, your mind is always working. Even during weekends, you are thinking about what may be happening on the floor.
I have lived that life. I have stayed at the office late, missed date nights, and missed events involving my children. That is the nature of the business.
With AI, I believe 70% to 80% of conversations can now be managed through automation in suitable use cases. That can substantially improve the lives of contact center leaders and employees.
It also changes how contact centers operate.
A large portion of repetitive work can be automated, while human agents can move into more meaningful, higher-value, and potentially higher-paying roles.
The designations, responsibilities, and overall operating model of the contact center will change.
Jake:
Can you give us a specific example?
What is one use case where AI has already delivered that kind of impact?
Ali:
One of the most interesting examples was also one of the most challenging.
The contact center supported brokers and enterprises with benefits administration and open enrollment.
The company had multiple clients, and each client had several categories of employees. Benefits differed from one person to another.
The system needed to understand details ranging from medical coverage to the level of parking access available when an employee arrived at the office.
This was during the earlier phase of our customer deployments. We expected it to be extremely difficult because of the number of systems, employee groups, benefit structures, and layers of information involved.
We continued working through each layer.
Within approximately two to three months, the AI was managing 84% of the calls, emails, and text-based interactions for that complicated use case.
That is the type of impact we are seeing from AI.
Not every use case reaches 84%. Some operate closer to 70%.
The lowest range we have observed has generally been approximately 70% to 72%. We have not yet seen many appropriate use cases fall below that range.
The main reason performance may be lower is that the client is using an old legacy system that cannot be integrated effectively.
An AI agent is only as capable as the data it can access.
In this case, the company had a relatively modern technology stack. We were also able to build agentic integration layers where direct connections were unavailable.
That helped us reach the final result.
Jake:
What happens to the remaining 15% to 30% of conversations?
Ali:
The main limitation is not always the technology itself.
There may be an exceptional or unfamiliar scenario that has not yet been provided to the AI.
The system may have access to the standard context, but a new situation appears that it has never encountered.
In that case, the AI may not be able to resolve the conversation immediately.
We then add that situation to the knowledge and training process, which allows performance to improve over time.
When the AI cannot resolve a conversation, it transfers the customer seamlessly to a human expert.
The human agent receives the context of the interaction and has the time needed to understand the issue, speak with the customer, and provide an appropriate resolution.
We also have an Agent Assist layer.
The AI can recommend the next action to the human employee. For example, it may recognize that a situation is highly unusual and advise the agent to speak with a supervisor.
The unresolved portion generally comes down to one of two things:
Jake:
You have said that your platform began outperforming much larger competitors. What do you think those companies were missing?
Ali:
I think they were missing context.
There are some excellent companies in the market.
At AM Infoweb, we were benchmarking multiple platforms because we wanted the best technology for our customers. At the same time, we were building our own platform.
We began to see that our results were stronger.
When I talk about results, my only North Star is conversation resolution.
I do not consider basic deflection alone to be a successful result.
We saw higher resolution rates, and we were surprised.
Many large companies have excellent technology teams, but those teams may not come from a contact center background.
Our team included contact center leaders and subject matter experts from different industries.
They understood the small operational nuances, what a real outcome looks like, and what contact center leaders actually need.
Initially, we repeatedly checked the results because we were comparing ourselves with multibillion-dollar companies.
We also used third-party reviews to make sure we were not overlooking anything.
The difference came from the context and operational knowledge brought by experienced contact center professionals.
Jake:
There are many voice AI systems entering the market.
Which companies do you think will survive, and which will struggle?
Ali:
There are approximately 3,000 companies in this category, which is an enormous number.
The difference between companies that survive, thrive, or disappear will not simply be their ability to build a platform.
Building a platform is becoming easier with vibe coding and other emerging technologies.
The more important questions are:
I can give you two examples.
Ali:
While we were building the conversational AI platform, another team in our organization was developing a medical records sorting, review, and summarization platform.
We came from that industry and had considerable domain knowledge.
The platform reached 86% accuracy, which was among the highest levels in the market at that time.
Other products were achieving approximately 70%, 75%, or 80% accuracy.
Some of those companies raised $15 million or $20 million during the early AI investment period.
Despite achieving better accuracy, we never launched our platform.
Because we had performed the work ourselves, we understood that 86% accuracy was not sufficient for medical records.
These records affect people’s health and lives.
The platform needed to reach human-level accuracy, which in that field was closer to 98% or 99%.
Without that level of accuracy, we did not consider the product usable.
Some competitors raised significant capital, spent the investment, and consumed the time of customers, but later struggled because the outcome did not meet the practical needs of the industry.
The same principle applies to conversational AI.
A company needs to understand what genuinely qualifies as a resolved conversation, what the technology should do, what it should not do, and where operational judgement is required.
Ali:
One of our customers was a healthcare provider.
They had approximately 40 or 50 people supporting their operation and asked whether AI agents could increase concurrent outbound calls from around 40 to 400.
Technically, the answer was yes.
However, making 400 calls simultaneously could flood payer lines and damage important relationships.
The organization wanted to get paid faster, but excessive calling could lead to delayed payments, strained relationships, or even legal and operational challenges.
The technology may allow you to scale, but that does not mean scaling without limits is the correct decision.
Someone with long-term experience in the industry understands those consequences.
That level of industry context, use-case understanding, and knowledge of how contact centers actually operate will determine which AI companies survive and grow.
Jake:
Do you see a future where AI replaces humans in the contact center?
Ali:
This is a question everyone asks.
Rather than giving a diplomatic answer, I will answer based on what I have observed over the last two decades.
Technology has generally benefited people when it is used responsibly.
Will AI replace humans in certain roles? Yes.
However, that does not mean those people will have no work.
I believe many employees will move into better, more satisfying, and higher-paying roles.
There may be challenges during the transition, particularly in the short term. Historically, however, new opportunities have emerged and overall consumption has increased.
Consider farming.
Approximately 100 years ago, farming was one of the largest occupations. Today, producing the same level of output may require only a small percentage of the effort once needed.
That did not eliminate work entirely. People found new industries and new things to do.
Humanity understands less than 1% of the universe. There will be many future roles connected to science, discovery, technology, and problems we have not yet imagined.
In the long term, I believe AI will create positive outcomes for people, even if the transition creates temporary challenges.
Jake:
Many AI companies are raising hundreds of millions or even billions of dollars.
You chose to bootstrap Orvera AI. Are you still committed to that approach?
Ali:
We have the luxury of not needing to raise money.
As the founder and CEO of another company, I have made mistakes and learned from them. One thing we managed reasonably well was financial discipline.
We do not carry significant debt, and from a financial perspective, we do not need outside capital.
That gives us the freedom to do what is right for the customer first.
Investors naturally need a return. They also have commitments to their own investors.
That creates pressure to show progress quarter after quarter.
Sometimes that pressure can cause companies to think too narrowly or make decisions that are not ideal for customers.
I have invested in other businesses and have seen the pressure experienced by executives. Through no fault of their own, they may need to meet targets that make it harder to prioritize long-term customer outcomes.
Bootstrapping gives us more freedom.
We would still consider outside capital if there were a strategic reason.
For example, if a partnership could expand our reach, accelerate adoption, and help more contact centers and leaders benefit from the platform, we would consider it.
However, we are not currently raising money because we need it.
Jake:
You have hired and managed thousands of people.
What advice would you give founders who are trying to move from 10 or 20 employees to the next stage of growth?
Ali:
This is one of my favorite topics.
There is no single answer because the strategy changes according to the size of the company, the location, and the stage of growth.
When we were a much smaller organization, it was difficult to attract highly experienced people.
Even when senior professionals from large companies joined us, they sometimes struggled because they were accustomed to very different operating environments.
Over time, I learned to focus on people with the right attitude, cultural alignment, willingness to learn, and ability to build alongside the company.
Eighteen years ago, I did not fully understand formal organizational culture and values.
However, I naturally attracted people who thought in similar ways.
Several leaders have now been with me for 15 to 17 years and continue to hold senior roles.
The original tagline for my first company was:
“We believe in relationships. Business follows.”
That principle continues across our companies.
Relationships work when every stakeholder benefits, people are treated with respect, and they feel part of the team.
Early in the journey, I focused on passionate people who were willing to learn.
As the company grew, we became able to attract more experienced talent and offer competitive compensation.
When a smaller company can provide meaningful equity to the right person, that can also be worthwhile.
At the current stage, I focus on hiring people who have recently completed the journey we are about to begin.
For example, if we are hiring an account executive, I may not choose someone from a multibillion-dollar corporation.
I may prefer someone from a company that completed our next phase of growth during the last one or two years.
That experience is more directly relevant.
A salesperson from a globally recognized company is accustomed to selling an established brand.
At a growing company such as Orvera AI, the salesperson must first build trust in the company and then sell the product.
Those are different skills.
Jake:
In Silicon Valley, we hear that companies increasingly look for cultural alignment, competence, curiosity, and the ability to use AI tools.
Has AI changed how you evaluate candidates?
Ali:
Not significantly, because we have always looked for those qualities.
Cultural alignment is essential.
In our organization, that means being ethical, thinking long term, and wanting to do what is right for the customer.
Competence is also essential, which is why we value experience relevant to the journey we are taking.
However, we have always avoided hiring people who behave as though they know everything.
Someone who believes they already know everything may not appreciate how quickly the world and the organization are changing.
A company that is growing more than 100% in a year is changing constantly.
In a services business, doubling growth may mean moving from 1,000 employees to 2,000 employees within a year.
That requires hiring, training, leadership development, and cultural alignment at a significant scale.
Our long-standing leaders have remained successful because they continually ask what has changed in the world, what has changed inside the company, and how they need to evolve.
AI is another major change, but that mindset was already part of the culture.
We also run workshops to help people understand the potential of AI, large language models, and the improvements these tools can bring to speed and quality.
Jake:
Voice AI adoption is still only a small percentage of the enterprise market.
What needs to change before it becomes mainstream?
Ali:
The technology is ready.
We see that through deployments with existing customers from our other business and through new Orvera AI customers.
The main limitation is that many leaders have not yet experienced the technology directly.
They may understand that AI is changing things, but they have not implemented it and therefore cannot measure the real impact.
There is a difference between intellectually understanding something and experiencing it personally.
I often compare it to having a daughter.
Before my daughter was born, I understood from watching my father that daughters are extremely dear to their fathers.
Actually having a daughter and experiencing that relationship is completely different.
AI adoption is similar.
Many leaders know that AI matters, but they have not yet experienced the operational result.
In some cases, the leader is ready, but the wider company is not.
The barriers may include senior leadership, technology stacks, compliance requirements, internal politics, and other organizational constraints.
A report I reviewed suggested that only 7% to 8% of enterprises have adopted voice AI.
There is usually a tipping point in an adoption curve. Once adoption reaches approximately 12% or 12.5%, momentum may increase significantly.
The technology is available. Adoption will take time because of the organizational challenges surrounding it.
Jake:
What are you most excited about over the next six, 12, or 24 months?
Ali:
We are growing aggressively.
Despite the challenges surrounding adoption, many companies are moving forward.
Healthcare has been particularly surprising.
Healthcare and financial services have traditionally been among the last industries to adopt new technology. However, healthcare is now one of our top target industries because adoption is moving quickly.
One reason may be the long-standing pressure to control costs.
The platform has already been stress-tested across multiple companies, use cases, and large volumes of daily conversations.
At the same time, we continue speaking with analysts, journalists, customers, and partners to identify smaller improvements.
I strongly believe in compounding. Small improvements can combine to create a significant advantage.
Ali:
Within the platform, customers have AI agents, 100% automated quality management, and Voice of Customer intelligence.
Previously, companies often needed to buy quality management and Voice of Customer capabilities from separate providers.
We include these capabilities with the AI agent platform.
Then we asked another question.
If the data is already available and AI is part of the platform, why should leaders still need to work through numerous charts and dashboards manually?
A leader should be able to ask a question in plain language, and the AI should analyze the data and provide an answer.
For example, imagine an ecommerce company delivering products across the United States.
A leader could ask:
“Which location has the highest number of delivery problems, and why?”
Instead of reviewing multiple dashboards and data points, the system could identify a specific ZIP code in New York and explain that refunds are higher because the logistics partner repeatedly fails in that area.
The company can then focus directly on fixing that problem.
Ali:
Contact centers were once viewed mainly as cost centers.
I believe they are now becoming command centers.
Organizations can conduct quality analysis across 100% of calls and capture the small details contained in the customer’s voice.
That is what we refer to as Voice of Customer.
Leaders can analyze information quickly without manually working through large volumes of charts and reports.
We are excited not only about the company’s growth, but more importantly about how these capabilities can help leaders and organizations make better decisions.
Jake:
Are there any roles you are currently hiring for?
Ali:
Our main focus is sales, particularly account executives in the United States.
During the last year, we stress-tested the platform, and the response from clients and partners has been very positive.
We are now investing more heavily in taking the platform to market.
Voice AI adoption is likely to increase significantly during the next few years, and we want to make the most of that opportunity.
We are also hiring operations professionals who support implementation after a customer comes onboard.
However, sales remains the main priority.
We already receive a meaningful number of inbound inquiries.
The account executive’s role is to speak with those companies, provide demonstrations, guide them through compliance and legal reviews, and support the full onboarding process.
That is currently the largest bottleneck.
Jake:
Ali, it has been great to finally have you on the podcast and learn more about your organization, your positioning, and the direction you are taking.
Where can people find you and Orvera AI?
Ali:
I am active on LinkedIn, and the company website is Orvera.ai.
People can connect with me through either channel.
Jake:
Thank you for joining us, Ali.
To everyone listening, thank you for spending your time with us.
I’m Jake Aaron Villarreal, signing off for now. We look forward to seeing you in the next episode.
