In May 2015, a 22-year-old with a 2.2 GPA posted a $9.69 Word document on Reddit. It wasn’t a pretty resume template — it was an argument about how resumes actually get read: not by a hiring manager first, but by software that most applicants never think about. Ten years later, that argument is a company called Rezi. No venture funding, ever. A Stripe-verified run rate north of $3 million a year. A content engine that out-ranks resume brands owned by the same conglomerate that owns Monster.com. And the whole thing is still run out of a small office in Seoul, South Korea, by a team you could fit around two tables.

Here’s how it actually happened, and the parts of it you can steal.

The bet nobody else was making

Jacob Jacquet’s resume got him interviews at Google and Goldman Sachs with a 2.2 GPA. That’s the origin story he tells, and it’s a fair one, because the mechanism behind it is the whole company: he’d researched how Applicant Tracking Systems — the software nearly every mid-size and large employer uses to parse, filter, and rank resumes before a human sees them — actually read a document, drawing on ATS research from Jobscan.co, and built his resume to survive that parsing pass rather than to look good on a desk.

In 2015, that was a genuinely different bet than the one everyone else in resume software was making. The category was, and largely still is, a design category — templates, colors, layout flexibility, one-click “professional look.” Rezi’s pitch was the opposite: strip the visual flourish, because a lot of it actively breaks ATS parsing. That’s still the explicit contrast Rezi draws against Zety today — Zety’s own comparison copy leans on multi-column, visually elaborate templates, while Rezi ships minimalist single-column layouts on the argument that a beautiful resume a scanner can’t read is worse than a plain one it can. In 2015, before “beat the ATS” was common job-search advice, that was a real, falsifiable claim about a specific mechanism — not a vibe.

Worth being honest about how rough the first version was: it wasn’t software. It was a Word document and a set of formatting rules, sold for $9.69 to $10, with no login, no database, no way to save your work in the cloud. That version shipped as a nights-and-weekends project while Jacquet was an undergrad, and it stayed roughly that primitive for almost four years. The gap between “a Word template a college kid is selling on Reddit” and “a real software company” is most of this story.

How the first customers actually showed up

Not “organic growth.” A Reddit post, specifically — Jacquet shared the ATS-focused template directly in a relevant subreddit, aimed at people wrestling with the exact problem he’d just solved for himself, and it moved by word of mouth from there among students in the same position he was in.

The second channel is more interesting and less repeatable-sounding on its face: on November 1, 2015, Jacquet moved to South Korea to teach English, which is how he funded the business — he wasn’t raising money, he was covering rent with a day job. He spent his first five months in Iksan, a small city with little startup infrastructure, before connecting with an English-speaking Korean collaborator who’d had San Francisco startup experience. Together they localized the product: Rezi.kr, a fully Korean-language version, launched March 27, 2016, and they partnered with JobKorea — a major Korean employment platform — to reach the large population of English-speaking Korean students and professionals applying to jobs abroad. South Korea sends one of the largest cohorts of international students to the United States, which made it a genuinely good-fit early market, not just a convenient one because Jacquet happened to be living there.

The moment that turned “guy running a side business” into “founder of a real company” was bureaucratic, not viral: in July 2017, Jacquet was accepted into the K-Startup Grand Challenge, a South Korean government program supporting foreign founders building in Korea. It gave him the legal and operational footing to formalize what had been, up to that point, an unofficial side project run alongside a teaching visa. If you’re looking for the highest-leverage “growth hack” in Rezi’s first two years, it isn’t a tactic — it’s a government accelerator program that let the founder keep building legally.

The growth machine

The quiet years

Rezi stayed a document, not a product, until September 2019, when it relaunched as a real cloud SaaS platform — accounts, authentication, a dashboard, the works. That’s also when Jacquet made his first hire, developer Luc Lemerez, who led engineering through the next two years of feature work: a resume-quality score, content analysis against ATS best practices, and AI-assisted keyword matching against job descriptions. None of this was flashy. It was the ordinary, unglamorous work of turning a validated idea into software people could actually use and pay for monthly. Jacquet’s own summary of that stretch, from his account of Rezi’s history: “Creating a business is a tireless process that will only cease when you give up or stop working completely.”

The GPT-3 bet

Here’s the part that reads like a lucky guess and wasn’t quite one. In December 2020, OpenAI’s GPT-3 was not a mainstream tool — it was closed, invite-and-application-gated technology; the waitlist for GPT-3 API access didn’t come down for the general public until November 18, 2021, almost a full year later. Rezi, then reportedly a team of around two people, got in during that gated window and shipped “Rezi AI Writer,” a GPT-3-powered feature that generated tailored resume content from a user’s job title. Contemporaneous tech press covered the launch in early 2021 — this wasn’t a claim manufactured after the fact once GPT mattered to everyone; it was reported while GPT-3 access was still a niche, developer-only story.

The bet wasn’t “AI is the future.” It was narrower and more useful than that: Rezi’s users had a specific, unsexy problem — most people are bad at describing their own accomplishments in resume language — and generative text was a plausible fix for that one problem, regardless of whether the broader world ever caught on. It happened to also be a two-year head start on brand association with AI-written resumes by the time ChatGPT launched in November 2022 and every job seeker on earth suddenly wanted “AI resume help.” Rezi’s own numbers show the tell: new customers were up 527.7% year-over-year in January 2023, the first full month after ChatGPT went mainstream. That’s not a company discovering AI. That’s a company that had already been living inside it for two years watching the rest of the market show up.

Turning on SEO

For years, Rezi had a good, ATS-differentiated product and, in the company’s own words to its SEO agency, no effective SEO strategy. That changed in late 2022, when Rezi brought on Skale, an SEO consultancy, for content and link-building work.

The first move wasn’t content at all — it was plumbing. Skale’s own case study credits a full rework of Rezi’s internal linking structure as the first material improvement, producing a visible jump in performance between January and February 2023, before a single new article shipped. Only after that did the content build start in earnest: topic clusters around resumes, cover letters, and resignation letters, each cluster reinforcing the others through internal links, plus custom landing pages tailored to different job sectors and seniority levels.

The specific, stealable mechanic sits inside that cluster strategy: templated pages built against a job-title matrix rather than one-off articles. Rezi didn’t write a single “resignation letter” post — it built resignation letter templates for dozens of individual job titles (content marketing manager, digital marketing manager, marketing analyst, and on), and did the same with resume examples by role. Each page is cheap to produce once the template exists, targets a specific long-tail search term a competitor’s generic page doesn’t, and links back into the core product. The flagship result: a resignation-letter-examples page pulled 109,000 non-branded clicks in a single month after optimization and link-building — a page about the act of quitting a job, ranking well enough to become a meaningful acquisition channel for a company that sells resume software. Rezi paired the content with affiliate partnerships across job boards and career sites, offering roughly 30% commission, to build the off-page link profile behind it.

Over the eighteen months Skale tracked (October 2022 to June 2024), non-branded clicks rose 292%, from roughly 23,000 to around 200,000 a month; organic sign-ups rose 243%; and revenue rose 86%. By Rezi’s own account, roughly a third of the top-ten Google results for “AI resume builder” belonged to Rezi pages in 2024 — a self-reported figure, but a directionally believable one given the traffic numbers Skale independently reported.

It’s worth naming exactly who that content engine was competing against, because “beat the big resume builders” undersells the size gap. Zety — founded a year after Rezi, in 2016 — is owned by Bold, a conglomerate whose portfolio also includes Monster.com, CareerBuilder, LiveCareer, MyPerfectResume, and FlexJobs. Resume.io, the other constant comparison point, was acquired by Career.io in 2021 and runs on the order of 50 employees. Rezi was doing this with a team you could count on two hands, against opponents with the marketing weight of a company that also owns Monster.com.

The honest version of this section has three engines, not one — the ATS-first product, the GPT-3 head start, and the SEO machine all compound together. But SEO is the one that did the most measurable, dated work between 2022 and 2024, and it’s the one a reader with no AI-lab access and no decade of runway can most directly copy. That’s why it gets top billing.

The enterprise pivot

By 2024, Rezi had enough individual-user proof to build outward instead of just upward. Rezi Enterprise, launched that year, is a white-label version of the platform that lets universities, workforce nonprofits, and recruiting organizations offer branded resume tools to the people they serve — priced around $99 a month per 200 job-seeker sign-ups, which Rezi pitches as roughly a $3 annual cost per user against a stated $29 for competing enterprise tools. By the end of 2025, more than 300 organizations had signed on, including named customers like Hofstra University’s career center and workforce nonprofit Merit America, plus at least one Fortune 500 company Rezi hasn’t named publicly.

The detail worth sitting with: Rezi’s first organic Enterprise subscriber — meaning inbound, not a sale their tiny team went out and closed — was Google, in February 2025. A bootstrapped, roughly dozen-person team in Seoul landed Google as a self-serve enterprise customer without an enterprise sales motion. That’s downstream of everything above it: the SEO traffic, the product credibility, and a decade of nobody having a reason to doubt the company was still around.

The numbers, plainly

Revenue by year, per Getlatka’s tracker (a third-party estimate, not an audit — cross-checked below): roughly $93,000 in 2020, $664,000 in 2021, $1.2 million in 2022, $2.4 million in 2023, and $3.2 million in 2024 (Getlatka’s own figure, flagged by them as an estimate). Rezi’s own year-in-review post is consistent with that trajectory without matching it exactly — the company reported crossing roughly $2.5 million in ARR by the third quarter of 2024, with January 2024 alone adding about $260,000 in ARR.

The strongest verification tier available for a private company is TrustMRR’s tracker, which pulls directly from a live Stripe API key rather than a self-reported figure. As of this writing (July 2026), that shows current MRR of roughly $259,000, trailing-30-day revenue of about $243,000, and cumulative all-time platform revenue over $9.6 million. Annualized, that current run rate is close to flat against 2024’s total — worth saying plainly rather than implying the growth line has kept climbing at the same slope it did through the ChatGPT-driven 2023 surge. It hasn’t; it’s plateaued at a real, sizable, still-fully-owned number, which is a different and more honest story than perpetual hockey-stick growth.

User growth, from Rezi’s own dated posts (a founder’s detailed disclosure, the strongest self-reported tier): 1 million total users by July 2023, with 65% of that first million joining in the preceding three months alone — the ChatGPT inflection, visible in the data. 2 million by the end of 2023. 3 million by November 2024. Just over 4 million by the end of 2025. That’s not a flat “about a million new users a year” — it’s an uneven curve, front-loaded by the AI wave, that has settled into something closer to a million-ish a year since.

Team size is the one figure I can’t fully reconcile. Rezi’s own year-in-review post puts the team at 18 people at the end of 2024 — 12 based in Seoul, 6 international across India, the UK, and Spain — which is the most specific, dated, primary-sourced figure available and the one I’d trust most. Third-party trackers disagree meaningfully: Getlatka shows 30 employees for 2024, and Tracxn shows 21 as of April 2026. I couldn’t find a way to square those with Rezi’s own count, so take the team size as “somewhere between a dozen and thirty people,” genuinely small relative to the revenue and traffic it produces, rather than a single precise number.

Funding: zero, independently confirmed by both Getlatka and TrustMRR. No institutional round, ever, in a decade of operation.

What you can actually steal — and what you can’t

Steal the wedge logic: pick a mechanism, not a feeling. “Passes ATS parsing” is testable. “Looks professional” is not. If your category’s incumbents are all competing on the same soft, unfalsifiable claim, there’s often a specific, provable one sitting right next to it that nobody’s built for directly.

Steal the content architecture, not just the idea of “doing SEO.” The mechanic that actually moved Rezi’s numbers was a template crossed with a long-tail keyword list — one job title at a time, one page at a time, each cheap to produce because the format was solved once. That’s available to you regardless of your budget; it’s a structuring decision, not a spend decision.

Steal the adoption logic on new technology, not the specific technology. Rezi didn’t win by predicting ChatGPT. It won by noticing that generative text solved an existing, narrow product problem its users already had, and building on it while it was still obscure enough that almost nobody else bothered. The transferable instinct is “does this new capability solve a problem my users already have,” not “get in early on whatever’s trending.”

Don’t expect to copy the Seoul setup directly. It’s real — Jacquet has pointed to South Korea’s technical, English-capable talent pool as a genuine advantage for hiring — but it’s downstream of a specific, personal set of circumstances: a founder who happened to move there to teach English, then happened to land a Korean government accelerator seat in 2017. That’s not a generic “go build your team in Seoul” playbook; it’s what happened when one person’s visa situation collided with a market that had no English-language resume-software leader yet.

And don’t expect to be first to the next GPT-3 moment by reading this article. That calendar slot is gone, for Rezi and for everyone else. What’s repeatable isn’t the timing — it’s the discipline of shipping on new capability because it solves a problem you already have, quietly, before anyone’s grading you on whether you called the hype cycle right.

Sources

From Rezi directly (founder/company disclosure):

Independent reporting and interviews:

Third-party trackers (cross-checked against each other and against Rezi’s own disclosures, not treated as sole sources):

Note on gaps: KoreaTechDesk’s full articles blocked direct automated access during research; the facts drawn from them here rely on search-indexed excerpts of the original reporting rather than the full text, and are corroborated by Rezi’s own accounts of the same events. Team size (18 per Rezi’s own count vs. 21–30 per third-party trackers) and the exact identity of Rezi’s Fortune 500 Enterprise client could not be independently confirmed beyond what’s stated above.

rezi seo bootstrapped content-marketing ai-tools
Share this article
LinkedIn (opens in new tab) X / Twitter (opens in new tab)
Atticus Li

Experimentation and growth leader. CXL-certified CRO practitioner, Mindworx-certified behavioral economist (1 of ~1,000 worldwide). 200+ A/B tests across energy, SaaS, fintech, e-commerce, and marketplace verticals.

About LinkedIn Newsletter