The AI Receptionist That's Generated $600M in Revenue for Clients
Source: No Priors | Published: 2026-07-31T10:00:08Z
Netic's AI agents answer customer calls and dispatch field jobs end-to-end, generating over $600M in revenue for clients. Founder Melisa Tokmak says using AI purely for cost-cutting would be a tragic waste.
Minus 20 Degrees, Nobody Picks Up
Picture this: late at night, minus 20 degrees, the heat suddenly goes out, and your kid is crying beside you. You find an HVAC company and call them — the voice that answers isn't human, but it knows your home type, your service history, whether a technician can come today, and which one to send.
That's what Netic does. Melisa Tokmak founded the company two years ago with a clear goal: help large enterprises in "essential services" — HVAC, plumbing, pet care, gyms, auto service — hand off the entire chain from customer contact to job dispatch to AI. She calls this the "autonomous enterprise." Companies focus on delivering great service. Netic handles everything else.
The Hard Part Isn't the Tech — It's the Dispatch Logic
On the surface, Netic is building "AI that answers the phone." The real complexity comes after.
When a customer calls to say their heat is out, the AI has to weigh several things at once: what's this customer's lifetime value? What does today's technician schedule look like? Does this need to be resolved today or can it wait until tomorrow? If the only available tech today specializes in boilers, is this customer worth burning that slot on?
Melisa says the operational complexity of these decisions is far beyond what most people imagine. Previously this work fell to hundreds of customer service reps and dispatchers — and the whole operation was inherently fragile. A company doing a billion dollars in revenue opens at 5 a.m., and on any given day three people might not show up and five might have quit — right as a heat wave sends calls through the roof.
More than 70% of Netic's current customers are already "N1 first" — meaning a customer's very first point of contact with the company goes directly to Netic's AI agent, not a human.
Why Build a Product Instead of a Roll-Up
The choice between building enterprise software and doing acquisitions was in front of Melisa before she even decided to start a company. She spent four years at Scale AI building out government, logistics, healthcare, and financial services business units. She understood the PE-backed, AI-transformation playbook perfectly — and she knew people were already running that play.
She chose the product route for three reasons.
First, skill fit. "I'm an engineer. I'm a product person. In an acquisition model, the most critical skill is M&A itself — and that's not what I'm good at. I didn't want to build a company where my most important strengths go unused."
Second, scale ceiling. A roll-up builds products that only serve the companies it buys. Nothing is reusable, nothing is replicable. "What I want is for every real-world business to be able to run on Netic."
Third, personal motivation. Melisa grew up in a small town in Turkey, with nearly every family member working in exactly these kinds of industries. She didn't have her own computer until she won a full scholarship to Stanford. She didn't want to build yet another tech company that only serves other tech companies.
The Foundation Models Aren't the Competition
"Ten years ago people asked if Google could do this. Now they ask if the AI labs can." Melisa's answer is direct: she doesn't see OpenAI or Anthropic as competitive threats — but the reasoning is worth unpacking.
Foundation model companies are optimizing for the most general solution possible. Melisa says if you put the problems Netic is solving in front of them, the answer would probably be: "Wait for AGI." She finds that intellectually and operationally lazy.
The other issue is focus. OpenAI ships products fast, but it also kills them fast — that's not a rhythm enterprise customers can live with. Anthropic found its footing by staying focused on coding agents, but on the enterprise side it's pushing roughly twenty products simultaneously, which creates confusion.
Meanwhile, Netic's customers speak with different accents, have different communication habits, and are in completely different emotional states — some calling in because it's the worst day of their lives. These "last mile" problems don't get solved at the model layer. They require enormous work at the product and orchestration layer.
Satellite Data for Door-to-Door Sales
Melisa finds one of Netic's customers — a large roofing company — genuinely interesting. The company employs door-to-door salespeople in the literal sense: they go block by block knocking on doors, pitching roof replacements or solar installations.
Now Netic pipes in satellite data: which neighborhoods took hurricane damage? What wear patterns are showing up across different roofing materials? That information gets fed automatically into the AI agent's context, so salespeople know exactly where to knock and how to open the conversation.
It's not that the company had never thought about this — they'd actually always wanted to use this kind of data. They just didn't have enough people to process it, and no single place to connect it with customer communications and work order dispatch. Netic puts the previously disconnected pieces on one platform.
This is also how Melisa pushes back on the assumption that these industries are conservative or tech-averse. She considers it a misconception. Some of the most tech-forward founders she's met work in the industries everyone assumes are old-fashioned.
PE Conversations Always Start With Cost
A large portion of Netic's customers are PE-backed businesses — some private equity firms hold twenty to thirty companies in this space. Melisa says the playbook has been shifting. The old approach was finding undervalued assets, swapping management, improving multiples, and selling. Those easy wins are increasingly scarce, so PE is now focused on actually creating value during the hold period.
But one pattern hasn't changed: the first conversation almost always starts with cost reduction.
Melisa's strategy is to redirect. "I'm not here to cut your costs. Cost savings will happen, but that's a side effect. I'm here to help you generate revenue you couldn't capture before." Then she pulls up a live deployment and shows actual customer data. To date, Netic's AI-handled interactions have generated over $600 million in revenue for its customers.
She says it would be a shame if AI were only ever used to cut costs.
The One Interview Question She Always Asks
Melisa has a fixed question in every interview: what's the hardest thing you've ever done in your life?
It doesn't have to be work-related. She says she's heard a wide range of answers, including from someone she recently hired. That person said: my life is simple, I'm very committed to my work, and I have a strict daily routine around my health. The hardest thing I've ever done is maintaining that routine for over fifteen years — every day, without getting bored, without getting distracted, staying focused on what I care about.
Melisa thought it was a great answer. She doesn't care how dramatic the story is. What she's looking for is: have you consistently demonstrated initiative somewhere in your life, and then seen it through?
She's deeply skeptical of the mindset that says "if you don't make it in 18 months you'll be permanently behind." In her observation, there's a kind of "permanent underclass" anxiety running through the younger generation — a sense that you have to learn everything in a very short window or AGI will leave you obsolete. She thinks that's dangerous. Doing one thing well takes a long time.
Education Is What Excites Her Most
When Melisa was in high school in Turkey, she could occasionally reach a few alumni who had gone abroad and ask them to look over her application essays or assignments — coordinating across time zones over Facebook Messenger. That was the full extent of resources she could access.
The AI application she's most excited about now is education. Not because the topic is fashionable, but because she lived through the specific pain of wanting to learn and having no way in. She believes AI is removing "access to resources" as a variable in the equation — what you want to learn, what you want to do, is becoming increasingly within reach.
But she adds one clear-eyed caveat: we're making choices easier, but most people still won't make them. Having or not having resources is no longer the constraint. Whether you act on it — that's still on you.