Monik Pamecha and Anthony Krivonos, Toma co-founders
Founder Profile — YesPress

Monik
Pamecha

The engineer who got himself banned from Reddit
trying to sell a diet app - and ended up in a car dealership in Oklahoma, listening to 4,000 phone calls, on his way to $17M from a16z.

AI Founder Voice AI Automotive Tech YC W24 a16z Portfolio
$17M Series A Raised
100+ Dealerships
138X ROI Q1
1M+ Call Minutes
$17M
Series A (a16z, Jun 2025)
100+
US Dealerships Active
45%
Calls Unanswered Industry-Wide
56hrs
Weekly Hours Saved Per Dealer
8-Fig
ARR Run Rate (Year 1)

Not Where You'd Expect a Silicon Valley Founder to Land

Picture Monik Pamecha in a service center in Oklahoma. It's 2024. He's not there for a tune-up. He's there because a car dealership owner called him up and said, "we are drowning in phone calls" - and Monik needed to hear exactly what that drowning sounded like before he wrote a single line of production code.

The path to that Oklahoma service bay started much earlier. Monik began coding around age 11 or 12, taught himself machine learning, published a guide to RNN-LSTM on Medium in 2016, and built one of the first conversational AI agents that same year. He didn't study computer science at a famous institution because someone told him to - he studied it because the alternative was going to waste.

By the time he arrived at Uber, then Turing.com (which he joined when it had about 10 employees and watched grow to a ~$4 billion company), then Braze, he had developed an unusually clear framework: "If there is any specific thing that you do that does not require creativity, it should be automated." Simple. Radical in practice.

His first real founding attempt after Berkeley was Skindex - an AI-powered skincare product discovery platform built as his MIMS capstone project - which was later acquired. Then came the startup that preceded Toma: a dietary management tool for patients with chronic conditions. The team got banned from Reddit and Facebook communities while trying to find users. Six months of nothing. So they made a Google Sheet.

Ten ideas. Two weeks each. The third or fourth: a horizontal voice AI platform. They prototyped it over a weekend and dropped it in a few online forums. Within three days, 200 demo calls were booked. Healthcare folks called. MLM people called. Construction companies called. And then - a car dealer called. And then that dealer recommended them to his "20 group."

"The car business chooses you," Monik said later. "You don't get to choose."

"If there is any specific thing that you do that does not require creativity, it should be automated."
Monik Pamecha — Co-Founder & CEO, Toma

Two Founders, One Thousand Dealerships Worth of Data

When Toma entered Y Combinator's Winter 2024 batch, the company was still small enough that Monik could - and did - spend most of his time not at a keyboard but in service bays. He sat in BDC offices in Oklahoma and Mississippi. He shadowed staff. He listened to how customers called about oil changes and how those calls somehow turned into missed appointments, lost revenue, and frazzled employees.

He and his co-founder Anthony Krivonos (formerly of Scale AI - yes, that Scale AI, the data-labeling company that became an AI infrastructure titan) described themselves as "plumbers" - connecting dealership management systems, customer databases, VIN records, and repair orders into a coherent picture that a voice AI could actually act on, not just talk around.

The product that emerged handles inbound calls, schedules service appointments in real-time, books test drives, processes parts orders, and follows up on no-shows - all while integrating with the existing software stack dealerships already run. It escalates when it needs to and stays out of the way when it doesn't.

The early product had a memorable bug: a cowboy-themed test prompt was accidentally left in the system, resulting in a Western-speaking AI greeting puzzled service customers. Monik talks about it now the way engineers talk about the bugs that aged into good stories.

The real validation came not from metrics decks but from dealers themselves. One called up after deploying Toma and said they were "drowning in phone calls" - pre-Toma, not post. That call confirmed exactly the problem Monik had been solving. By the time a16z partner Seema Amble led the $17 million Series A in June 2025, Toma was already handling over 1 million call minutes and had crossed seven-figure ARR in under a year. The a16z team's description of Monik and Anthony's fieldwork: "They were effectively living at these dealerships."

Dealerships using Toma report saving 30 to 56 hours per week. Some report 138X return on investment in the first quarter. In an industry where BDC employee turnover runs at roughly double the national average, an AI that doesn't quit has uncommon staying power.

Toma at a Glance

📞
1M+
AI Call Minutes
Since Launch
🏪
100+
Active U.S.
Dealerships
💰
$17M
Series A
(a16z, Jun 2025)
138X
Reported ROI
in Q1
🕐
56 hrs
Weekly Hours
Saved (Max)
📈
<1yr
To 7-Figure
ARR
🚗
$2T
US Auto Industry
Market Size
📊
45%
Dealership Calls
That Go Unanswered

What Monik Actually Says

"The car business chooses you. You don't get to choose."

"Voice AI is just a channel - it's really the work that gets done behind the receiver that's complicated."

"Trust is created when customers' needs are addressed quickly and accurately."

"When was the last time something in your life went according to plan?"

"The things that you have to transfer for will keep shrinking over time - as AI handles increasingly complex scenarios."

"We built AI agents that don't miss a beat and provide an exceptional customer experience while getting more intelligent."

"They were effectively living at these dealerships."
Andreessen Horowitz Partner — on Monik Pamecha & Anthony Krivonos's field research approach

From Mumbai to the Service Bay

~2012
Begins coding around age 11-12 in Mumbai, India.
2016
Publishes ML research on generative models for chatbots - one of the earliest conversational AI agents. Starts technology blog Etiole.com at 17; it peaks at 100,000+ monthly visitors and hits Digg's front page.
2016 - 2017
Works at IIIT Hyderabad and Haptik Inc (conversational AI). Writes AI/ML tutorials on Medium.
2017 - 2019
Masters (MIMS) at UC Berkeley School of Information, specializing in Data Science.
2019 - 2020
Co-founds Skindex - an AI skincare product discovery platform built as his Berkeley capstone. The company is later acquired.
2020 - 2022
Early employee at Turing.com (~10 people when he joins; the company eventually reaches a ~$4B valuation). Also works at Uber and Revcontent.
2022 - 2023
Joins Braze (marketing technology). Begins exploring new startup ideas.
2023
Attempts a dietary management tool for chronic condition patients. Gets banned from Reddit and Facebook communities. Pivots. Makes a Google Sheet with 10 ideas.
Jan 2024
Joins Y Combinator Winter 2024 batch. Prototypes a voice AI platform over a weekend; 200 demo calls booked within 3 days. Automotive dealerships emerge as the strongest market fit.
2024
Spends months on-site at dealerships across Oklahoma and Mississippi - sitting in BDC offices, shadowing staff, listening to calls at 3x-4x speed.
Jun 2025
Raises $17M Series A led by Andreessen Horowitz (partner: Seema Amble). Toma active in 100+ U.S. dealerships, approaching 8-figure ARR. Announces on X: 45% of dealership calls go unanswered - Toma fixed that.
Feb 2026
Presents at NADA Show 2026, interviewed by CBT News. Continues expanding Toma's platform capabilities.

Things That Actually Happened

Field Work

When Monik and Anthony arrived at dealerships in Oklahoma and Mississippi to train their AI, they weren't running workshops. They were listening - sitting in service bays and BDC rooms for days at a stretch. Monik is vegetarian. When the dealership owners cooked home meals during those visits, things got "awkward but funny." He tells it as a good story.

The Reddit Ban

Before Toma, Monik and his co-founder spent six months on a dietary management app for chronic condition patients. They tried to seed the community through Reddit and Facebook health groups. They got banned from both. The product never found its users. That failure taught them something: product-market fit shows up fast when it's real. It was fast with Toma.

The Google Sheet

After the diet app failed, they built a list of 10 ideas and agreed to spend exactly two weeks testing each one. No falling in love. No sunk-cost logic. The third or fourth idea on the list was a horizontal voice AI platform. They built a weekend prototype. Posted it in forums. Three days later: 200 demo calls scheduled. The spreadsheet never made it past row four.

The Cowboy Bug

Early in Toma's development, an engineer left a test prompt in the system - a cowboy-themed persona for a Friday demo. Forgot to remove it. The next day, a service customer called a dealership and was greeted by an AI speaking in a distinctly Western twang. Monik tells this story with the specific fondness of someone who has survived enough launches to know which bugs become mythology.

4,000 Calls

Monik has personally listened to more than 4,000 dealership phone calls - often at 3x or 4x speed. Not to audit the AI. To understand the patterns. He found that the vast majority of service calls follow predictable if-then logic: same questions, same objections, same stall points. That insight became the architectural foundation of Toma's resolution engine.

Automating Lunch

The "automate what doesn't need creativity" principle isn't just a pitch line. Monik uses AI to randomize his DoorDash orders. He's serious about this. It started as an experiment and stayed because it worked. The founder of an AI voice platform for car dealerships does not choose his own dinner. The model does.

What He's Actually Built

Raised $17M Series A led by Andreessen Horowitz - in June 2025, barely 18 months after entering YC
Scaled Toma to 100+ U.S. dealerships and 8-figure ARR run rate in under two years
Built one of the earliest conversational AI agents in 2016 - a full eight years before Toma
Co-founded Skindex (AI skincare platform, Berkeley MIMS capstone) - later acquired
Joined Turing.com when it had ~10 employees; the company later reached a ~$4B valuation
Technology blog Etiole.com reached 100,000+ monthly visitors and was featured on Digg's front page
Toma handles 1 million+ call minutes since launch; dealerships report 138X ROI in their first quarter
Featured speaker at NADA Show 2026; named a judge for the AI Poverty Challenge for tech-driven social good

What Doesn't Make the Press Release

Real Madrid fan. In a room full of Silicon Valley founders betting on AI to disrupt everything, Monik is betting on Los Blancos.

🥗

Vegetarian - which, when you're living inside Midwestern car dealerships for weeks at a stretch, creates genuinely awkward situations with home-cooked meals.

🎙️

Has listened to over 4,000 dealership phone calls, many at 3x-4x speed. If there is a sport for this, he is a contender.

🍔

Uses AI to randomize his own food delivery orders. The CEO of an AI automation company does not choose his own lunch. Practice what you automate.

📝

Published one of the earliest guides to RNN-LSTM in TensorFlow on Medium in 2016 - back when most ML practitioners were still reading research papers and guessing.

🤝

Committed to technology as a tool against poverty - serves as a judge for the AI Poverty Challenge and previously organized Twestival, which raised millions for charity:water.

🐄

His co-founder Anthony Krivonos sold nipple covers on Amazon before Toma. Monik has mentioned this publicly. In at least one version of this story, it's context for why neither of them was afraid of an unusual market.

🤖

Advocates always disclosing to callers when they're talking to an AI. A Toma agent that hid its identity generated calls asking "Are you human?" repeatedly. Transparency is also product design.

Monik in Conversation

Monik Pamecha - Toma
An introduction to Monik and Toma's vision for AI in the automotive dealership industry.
Listening To 4,000 AI Phone Calls
Auto Collabs podcast: Monik digs into what he learned from analyzing thousands of dealership calls and how it shaped Toma.
Meet the Mastermind Behind Toma's AI
A closer look at Monik's background, the founding story of Toma, and the technology driving the platform.
Podcast
Giant Robots: How Toma is Transforming Car Dealerships
ThoughtBot episode 534 - the full story of Toma's origin and Monik's approach to building voice AI in an overlooked industry.
Listen to Episode
Share This Profile