How Pavilion Hit $10m in Revenue While Balancing Investor and Customer Demands with CEO Sam Jacobs

Jonathan Adly spent 15 years as a licensed pharmacist before writing the first line of code for TJM Labs in November 2024 — and 18 months later he's at $25M ARR, 40% EBITDA, 450 pharmacies, and $100M raised without ever making a pitch deck. His pharmacy AI agents cost $45,000 a year, replace 2–3 human technicians each, and now fill 500,000 prescriptions every day. In this episode he breaks down the $1M → $15M → $25M ARR run, why he took a $75M Series B from Elephant while already profitable, how he values an AI company at 15X+ off public comps, and the napkin math from 450 pharmacies to a $1.8B opportunity.  

Vadim Yasinovsky, co-founder of pdfFiller and CEO of airSlate (parent of pdfFiller, signNow, DocHub, US Legal Forms, airSlate Flow), breaks down how a plain-looking PDF editor page became a $106M/year product with 866,000 paying customers — inside a company doing $200M in revenue across 1.1 million customers at an average of $180/year. He raised just $150M total (last round: $52M at a $1.25B valuation in June 2022) and now says the company is optimistically worth $800M — despite doubling revenue since. He bought signNow at $3M and grew it to $35M. He built an SEO machine on 4.8 billion keywords and 35 million pages, and now says that playbook is dead within 9–12 months as ChatGPT and AI search take over. In this episode: Why 650 employees are split into siloed 120-person product teams with no shared tech Why airSlate didn't become Harvey despite 85,000 lawyer-built legal forms as a training set Why Vadim says the "ChatGPT SEO" gurus are "b.s." and his influence on search drops from 80% to 30% His unannounced AI product (bizdraft.com) and why he won't monetize it The path from $200M to $400M: 20–30% from existing products, the rest from vertical AI His M&A filter: product-market fit, domain expertise, $5–10M+ revenue minimum His advice to founders: raise as late as you can, cash is king, preserve it now Topics: SaaS, SEO, programmatic SEO, AI search, ChatGPT, document automation, e-signature, PDF editor, legal forms, SMB software, bootstrapping, vertical SaaS, M&A, capital efficiency.

Jeff Lu quit Google Cloud in 2022 to build his own AI video models at Akool, a generative AI platform for avatars, face swap, and video translation. He skipped the giant foundation models and trained small ones for specific tasks on rented servers, then watched revenue go from $100K in 2022 to $3M in 2023, $13M in 2024, $25M in 2025, and over $3M in June 2026 alone. He breaks down why Coca-Cola picked him over the foundation models and now pays him $4m per year. He'll share how he's driven huge usage (500,000 images and videos created daily) and what he'd say if Google offered $500M cash. Akool revenue data: https://getlatka.com/companies/akool.com

Antoine Fort launched Qobra in 2020 with a $100K pre-seed from 7 angels, and his first investor was also his first customer. He started by cold calling with a phone list from a provider later shut down over European privacy law (GDPR), charging $10K/year. Today Qobra is at about $30K/year per customer, 250+ customers, and roughly $7–8M ARR. Along the way, one meme ad closed an $80K/year deal, and a single sales rep on the platform earned $1.5M in commissions (75x the average). So when Nathan asks if he'd sell for $100M cash today, his answer is surprisingly honest. Qobra is an AI-first sales compensation and commission software platform. Antoine covers his $5M seed and $10M Series A, the $2M in debt he raised to limit dilution, the $200K–300K/year he spends on LinkedIn, YouTube and Google ads, his G2 and SEO glossary playbook, and the three AI agents Qobra is building to replace commission consultants.

Sevan Marian, co-founder and CEO of Fleet, bootstrapped his Paris-based laptop rental company for 7 straight years to roughly $40M in annualized revenue and close to $10M in yearly profit (EBITDA), without raising a single dollar of VC. His trick: customers pay about $50/month on 36-month contracts, and Fleet sells each contract to a bank that pays it upfront on day one. That's how he's financed 50,000+ devices across 20 countries without holding inventory. In early 2026 he sold about 25% of the company at a $100M valuation, and every dollar went out to the team: employees cashed out 100% of their shares, worth several million euros.  

At 23, Alex Jekowsky started building Cents, an all-in-one software, hardware, and payments platform for laundromats, dry cleaners, and shared laundry rooms. In 2025, Cents generated $60 million in revenue, up 67%, with just 80 employees and EBITDA margins above 20%. The company now has 99% customer retention, powers laundry rooms in nearly 7,000 apartment buildings, and recently raised a $140 million Series C at an approximately $800 million valuation. In this episode, Alex reveals: • How Cents grew from $700,000 to $60 million in four years • Why its first customer was a 73-location laundry operator • How Cents combines software, payments, and custom-built hardware • Why the company generates roughly $750,000 in revenue per employee • How Cents maintains 99% customer retention • Why Alex acquired a hardware company instead of continuing to build everything internally • How the $140 million financing included a $30 million employee tender • Why founders should avoid raising at valuations that eliminate future options • How Alex plans to turn an overlooked industry into a multibillion-dollar company

Froda generated $57 million in revenue in 2025 by lending to the small businesses traditional banks often overlook. In this episode, Nathan Latka sits down with Froda co-founder Oliver Mohseni to unpack the economics behind the startup bank. Froda has supported 150,000 businesses across seven markets and funded approximately $3.45 billion since launch. Oliver explains how Froda uses retail deposits, automated underwriting, and embedded partnerships with banks, fintech companies, payment providers, and software platforms to scale its lending operation. Listen to learn: • How Froda makes money on an average loan of just $23,000 • Why its banking license and retail deposits create an advantage • How it prices loans at approximately 16% APR while managing credit losses • Why embedded lending helped Froda expand across Europe • How the company built a roughly $500 million active loan book

Carl Turner, founder of Swipe.by, built an AI marketing platform for restaurants to $3.5M ARR growing 400% year-over-year — with $0 spent on ads and every customer closed by a rep knocking on doors. He breaks down the exact rep comp plan (6 locations/month, $6K ACV, 30–40% commission), why he took $450K from Founderpath instead of Stripe, and why he gives away websites and online ordering for free. Then he turns down a $15M cash offer and explains the 2x liquidation preference math that could leave a founder with nothing. SaaS, bootstrapped, AI, restaurant tech, outside sales, SMB, Nathan Latka, Top Founders

Raphaël Boukris is the co-founder and Chief Revenue Officer of Didomi (didomi.io), the Paris-based consent management platform that helps enterprises comply with GDPR, CCPA, and privacy laws worldwide. Roughly 2% of all global web traffic passes through Didomi's SDK, and the company processes billions of consent decisions every month for customers including Bloomberg, Sky, and Adyen. In this interview, Raphaël breaks down the full financial history on the record: three funding rounds, two acquisitions, and a majority sale to private equity. He raised under €300k from angels, then €5m from Breega in late 2019, then a €34m ($40m) Series B led by Elephant and Breega in summer 2021 — priced, he says, at above a 10x ARR multiple. Eight months later the market collapsed. He never burned through the money. In April 2025, Marlin Equity Partners took majority control, and Didomi immediately acquired Addingwell and Sourcepoint. Today the group does $40–60m ARR across 3,500 customers with fewer than 200 employees, and has at least one customer paying over $1m per year. What we cover: (01:02) What Didomi actually sells, and why every cookie banner in Europe is a business (01:51) Why roughly 2% of global web traffic runs through their SDK (02:12) Processing billions of consents per month — and who is not the customer (03:00) Pricing on monthly unique visitors, and why AI agents and bots are breaking that model (04:32) The Addingwell acquisition: a few million in ARR, a few thousand euros ACV, profitable enough to pay dividends (05:10) Why he bought a second product instead of building one (06:56) Every round, in order: sub-€300k angel, €5m Breega, €34m Series B (07:50) Above a 10x multiple in 2021 — and how he ran Elephant through a roadshow to get it (09:44) Secondaries in the Series B, and managing a team whose options went underwater (10:06) The year he crossed $1m ARR, and the year he crossed $10m (12:18) The question he answers "absolutely not" — was he ever close to shutting it down (13:09) What growth equity actually pays in 2025, and the range he'd give (14:22) Buying Sourcepoint, a company that had raised $48m, for its ACV and its US enterprise base (15:07) 3,500 customers, one paying over $1m/year, under 200 people (16:44) The combined group revenue number: $40–60m ARR Full metrics, funding history, and revenue timeline: https://getlatka.com/companies/didomi.io  

Stéphane Donzé has bootstrapped AODocs and Talarian to a combined $60m in 2026 revenue without raising capital. His top three customers each pay more than $2m per year, including Google, which pays him millions annually to control the documents behind its data center builds. He funded fourteen years of R&D off a cloud consulting business, never burned more than $200k in a month, and has fielded private equity calls four times a week declining $600m offers.