Karvi: Revenue, team and how it was built
Karvi (AI review of LinkedIn profiles, CVs written for one job ad, and interview preparation for job seekers, Brazil): US$8.6K all-time Stripe revenue, November 2026. Evidence grade A: disclosed by a company or a filing, or verified by a payment processor. Team: 1 person (founder Bruno Bertolini), as stated on the TrustMRR profile. Launched February 2026; still running, revenue falling.Added 6 November 2026
- Outcome
- US$8.6K all-time Stripe revenue, November 2026Revenue falling
- Evidence
- A disclosed by a company or a filing, or verified by a payment processor
- Launch
- February 2026
- Status
- Still running
- Team
- 1 person (founder Bruno Bertolini), as stated on the TrustMRR profile
- Outside funding
- None (bootstrapped, founder-provided on TrustMRR)
- Main channel
- Founder's posts and audience
Karvi is a set of AI tools for job seekers, built by Bruno Bertolini, a software engineer in Brazil. Its first tool is a LinkedIn profile analyzer: a program that reads the text of a LinkedIn profile, compares each section with rules for what recruiters search for, gives a score, and writes improved text. The user uploads the PDF that LinkedIn itself exports, so Karvi does not need to read LinkedIn pages. Later tools write a CV for one job ad, prepare interview answers and track job applications.
Karvi sells credits instead of a subscription. A credit is a unit of prepaid usage: one LinkedIn analysis costs 100 credits, one CV for a job ad costs 20. The Starter pack of 100 credits costs R$19.90 in Brazil (US$3.90 on the English page), and the Full pack of 1,000 credits costs R$99.90 (US$19.90). Credits do not expire. TrustMRR, a site that reads revenue directly from a company's payment provider, shows US$8,590 of Stripe revenue from January to 6 November 2026, with US$1,868 in February 2026, the first full month.
The Stripe figure is not the whole revenue. The founder wrote on the TrustMRR listing that 83% of the revenue is paid by Pix through Mercado Pago and Woovi, which TrustMRR cannot read, and that lifetime revenue is about US$49,800; that figure is his claim only. Both sources agree that sales fell after July 2026: Stripe revenue was US$787 in August and US$18.33 in October. Since 2 September 2026 the business has been listed for sale on TrustMRR with an asking price of US$50,000.

How much revenue does Karvi make?
Each row shows the evidence grade and the source numbers from the list at the end of this page. The grades are explained in How Aiolum verifies.
| Fact | Value | Grade | How we know | Sources |
|---|---|---|---|---|
| Verified Stripe revenue, all time to 6 November 2026 | US$8,590 | A | TrustMRR, read from the Stripe account through an API key; revenue last synced 6 November 2026. The last-12-months figure is the same, US$8,590. | [1] |
| Stripe revenue, February 2026 (first full month) | US$1,868 | A | TrustMRR monthly timeline. January 2026 shows US$5.74. | [1] |
| Stripe revenue, best months after launch | US$1,393 (May) and US$1,456 (June 2026) | A | TrustMRR monthly timeline. March US$684, April US$1,151, July US$1,132. | [1] |
| Average Stripe revenue, February to July 2026 | about US$1,281 a month | calc | Calculated: (1,868 + 684 + 1,151 + 1,393 + 1,456 + 1,132) divided by 6 equals 1,280.67. | [1] |
| Stripe revenue, August to October 2026 | US$787, US$88.49, US$18.33 | A | TrustMRR monthly timeline. October is 99% below February (calculated: (1,868 - 18.33) / 1,868). | [1] |
| Current monthly recurring revenue | US$0 | A | TrustMRR shows US$0 MRR and 0 active subscriptions. This is expected: Karvi sells one-time credit packs, not subscriptions. | [1, 3] |
| Lifetime revenue including Pix payments | about US$49,800 across 13,605 paid transactions | C | The founder's message on the TrustMRR listing: "83% of revenue is PIX via MercadoPago and Woovi, which TrustMRR can't read". Not verified; he offers gateway exports to buyers under a non-disclosure agreement. | [1] |
| Average payment in the founder's claim | about US$3.66 | calc | Calculated: 49,800 divided by 13,605 equals 3.66. This is close to the US$3.90 Starter price. | [1, 3] |
| Stripe share of the founder's claimed total | about 17% | calc | Calculated: 8,590 divided by 49,800. It agrees with the founder's statement that 83% of revenue is Pix. | [1] |
| Founded | 3 February 2026 | C | Founded date on the TrustMRR profile, entered by the founder. | [1] |
| Team | 1 person | C | TrustMRR profile, founder-provided. No source suggests more people. | [1] |
| Outside funding | none | C | TrustMRR profile: "Funding: Bootstrapped" (founder-provided). | [1] |
| People who ran a profile through Karvi | 52,853 (July 2026); 71,523 (November 2026) | C | Counter on the home page: "+ 52.853 profissionais já passaram um perfil pela Karvi" in the Wayback copy of 10 July 2026 and "+ 71.523" on 6 November 2026 (translation: professionals have already run a profile through Karvi). Most of them use only the free diagnosis. | [2, 5] |
| Profiles analyzed, January to May 2026 | 26,507 | C | Karvi's own report "State of LinkedIn in Brazil in 2026". 97.6% of the profiles with a detected language were in Portuguese. | [6] |
| Visitors, last 30 days | 2,724 | A | TrustMRR, from the connected Google Analytics account. | [1] |
| For-sale listing | asking US$50,000, listed 2 September 2026 | A | TrustMRR acquisition section. The listing reports 0 offers. | [1] |
How to read this table: a grade A value was read by TrustMRR from the company's Stripe account. A grade C value is a founder statement or a figure from the company's own pages. A "calc" value (grade X) was computed on this page from the rows above it.
| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| First diagnosis | 0 | 0 | 1 free LinkedIn diagnosis, no card |
| Single pack | 1.90 one-time | 1.90 one-time | For one action; credit amount not shown on the page |
| Starter | 3.90 one-time | 3.90 one-time | 100 credits: 1 LinkedIn analysis or 5 job analyses with CV |
| Full | 19.90 one-time | 19.90 one-time | 1,000 credits: 10 LinkedIn analyses or 50 job analyses with CV |
Brazilian prices on the Portuguese page: Starter R$19.90, Full R$99.90, single pack R$7.90, paid by Pix or card. The table gives the US dollar prices from the English page. What credits buy: LinkedIn analysis 100, CV for a job ad (analysis included) 20, job search 20, interview preparation 25, CV regeneration 5, AI professional photo 20. Credits never expire. The first LinkedIn diagnosis is free and needs no card.
When did Karvi launch, and how did it grow?
From launch in February 2026 to a for-sale listing seven months later. Each entry carries its grade.
- January 2026AFirst Stripe payment, US$5.74, before the official founding date. [1]
- 3 February 2026CFounding date on the TrustMRR profile. The LinkedIn analyzer is the first tool. [1]
- February 2026AUS$1,868 of Stripe revenue in the first full month, the best month so far. [1]
- January to May 2026CKarvi analyzes 26,507 LinkedIn profiles, almost all in Portuguese, and later publishes the aggregate data as a report. [6]
- May to June 2026AStripe revenue recovers to US$1,393 and US$1,456 a month. [1]
- 10 July 2026CThe home page counts 52,853 professionals and sells four tools: LinkedIn analyzer, CV per job ad, interview preparation and an application tracker. [5]
- August to October 2026AStripe revenue falls to US$787, then US$88.49, then US$18.33 a month. [1]
- 2 September 2026AThe business is listed for sale on TrustMRR at US$50,000. The founder writes that real lifetime revenue, with Pix, is about US$49,800. [1]
- 6 November 2026ACredit packs from US$1.90; the home page counts 71,523 professionals; US$8,590 of verified Stripe revenue in total. [1, 2, 3]
Why did Karvi work?
Four causes, in the order they mattered. Each one is a reusable lesson.
A free diagnosis that shows the problem
The first analysis is free and shows a score and the weak sections, for example "Headline sem palavras-chave" (translation: headline without keywords). The rewritten texts are paid. The free result does the selling, and the price is low enough to pay on impulse.
Pay once, in the local payment system
There is no subscription and no trial, and credits never expire. The founder said on TrustMRR that 83% of payments are Pix, Brazil's instant bank payment. A one-time Pix payment of R$19.90 is a small decision for a person who is looking for a job right now.
Input the platform already exports
The user downloads the PDF that LinkedIn makes with its own "Save as PDF" button. Karvi reads that file, so it does not depend on reading LinkedIn pages or on LinkedIn's interfaces.
Calibration by profession
The home page says each analysis is calibrated by profession, level and field, and the site has requirement pages for careers such as design, backend, data analyst and DevOps. This is the difference it claims over asking ChatGPT directly.
How it was built
Karvi has not published a technical write-up. These details come from the founder-provided tech stack on TrustMRR and from Karvi's own pages (grade C). [1, 2, 4, 6, 7]
- Next.js for the web app, hosted on Vercel.
- Neon, a hosted Postgres database, for accounts, credits and analyses.
- Resend for email, and Stripe for card payments outside Brazil. The founder says Mercado Pago and Woovi process the Pix payments.
- Sign-in with LinkedIn, Google or GitHub, or by an emailed link (sign-in page).
- Input is the PDF that LinkedIn exports. The analysis covers more than 30 criteria in 6 areas and returns texts ready to copy. No source names the language model behind it.
- Content pages for search: guides to the LinkedIn headline, about, experience and skills sections, a page on LinkedIn's 360Brew ranking model, career requirement pages, and a data report built from 26,507 analyses.
Tools named in the sources
Each one opens its page in the Tools tab, with every story that uses it.
Who tried this before Karvi, and why did they fail?
Three companies in our database of failed startups sold help to job seekers before language models could write a profile or CV for one person. The values below are from that database and are not independently verified.
Jobspice
The promiseProfessional-looking resumes made quickly from templates.
Why it diedIt offered only resume building. Free templates spread online, and LinkedIn combined profiles, networking and job search in one place.
LearnUp
The promiseJob training modules that prepare candidates for specific employers.
Why it diedCustom content and employer partnerships were expensive, and LinkedIn Learning and Coursera offered more courses with larger networks.
Jobr
The promiseA mobile app to swipe through jobs and apply quickly.
Why it diedLinkedIn and Indeed had more jobs and more users, and the swiping idea did not create a lasting advantage or a way to make money.
The failed startups come from our database of 1,749 dead startups, retold here in our own words. Its entries are AI-assisted summaries of public sources, so these cards carry no evidence grade.
What changed
Before language models, help with a profile or a CV meant a template or a person. Jobspice sold templates, and free templates replaced it. A career consultant writes text for one person but costs much more: Karvi's own comparison table lists R$400 to R$1,500 per project for a human consultant.
A language model can read one person's profile and write new text for that person in about two minutes, at a cost low enough to sell for a few dollars. Karvi adds the parts around the model: rules per section, calibration by profession, a score, and a free first result. The low price and the Pix payment let one founder sell it to a mass market in Brazil.
Can one person copy Karvi now?
Yes, the method is simple to copy; the hard part is keeping sales after the launch.
One person built it with common tools (Next.js, Vercel, a hosted Postgres database, Stripe). The site already lists several similar stories, for example Rezi, Jobright and Resumeaze. Karvi's own Stripe data shows the risk: the first month was the best month, and revenue in October 2026 was 99% lower.
Where it is still open: The same tool in other markets with their own language and payment habits, and versions for one profession where the rules are specific (nurses, lawyers, engineers who must show licenses).
- 1Pick one document that people are judged by (a LinkedIn profile, a CV) and a format they can export themselves.
- 2Give a free score that names the weak parts, and charge a small one-time price for the rewritten text.
- 3Accept the payment method your market uses most: in Brazil that is Pix.
- 4Write guide pages and a data report from your own aggregate data, so people find you after the launch posts stop.
What could go wrong for Karvi?
The risks are the mirror image of the causes above.
Falling revenue
Stripe revenue was US$1,868 in February 2026 and stayed above US$1,000 in most months to July, then fell to US$787 in August, US$88.49 in September and US$18.33 in October (TrustMRR). The founder says most revenue is Pix, which TrustMRR cannot see, so the Pix trend is not known.
Revenue outside Stripe is not verified
The founder's claim of about US$49,800 in lifetime revenue and 13,605 paid transactions rests on his own message on the TrustMRR listing. He offers gateway exports under a non-disclosure agreement to buyers; no public source confirms them.
Listed for sale
Since 2 September 2026 Karvi has been listed on TrustMRR with an asking price of US$50,000 and, at reading, 0 offers. A new owner may change the product or the prices.
Outcome claims on the sign-in page
The sign-in page shows "+1,950% visualizações de perfil em 1 dia" and "2 propostas de emprego em 2 dias" (translation: +1,950% profile views in 1 day; 2 job offers in 2 days) without a source or a name. The home page itself says the texts are writing guidance, "não uma previsão de posição na busca ou contato de recrutadores" (translation: not a prediction of search position or recruiter contact), and that the AI can make mistakes. The testimonials on the home page carry full names and say they are screenshots of real messages; this page did not check them.
Competition and model providers
The site already has several jobs-and-careers stories with similar tools, such as Rezi, Jobright and Resumeaze. No Karvi page names the language model it uses; its FAQ only compares the product with ChatGPT, so the dependence on one model provider and its prices cannot be judged from the sources.
Other product of the same founder
The founder's TrustMRR name links to awesome.codar.me, which presents an AI landing-page analyzer called AwesomeLanding. It is not part of this story.
Sources
The grade next to each source is the best grade that the source can support. "Read in full" means that the text of the page was downloaded and the numbers were checked in it.
- A TrustMRR profile: Karvi (revenue synced 6 November 2026)Read the markdown profile (https://trustmrr.com/startup/karvi.md) in full on 6 November 2026. Revenue is read from Stripe through an API key (grade A); the founding date, team size, funding, tech stack and the seller's message are founder-provided (grade C).
- C Karvi home page: Karvi · Inteligência de Carreira por IARead in full on 6 November 2026 (Portuguese). Source for the tools, the user counter, the testimonials, the FAQ and the comparison with consultants.
- C Karvi pricing pages (Preços and Pricing)Read the Portuguese page and the English page (https://karvi.app/en/pricing) in full on 6 November 2026. Source for the credit packs and what credits buy.
- C Karvi LinkedIn Profile Analyzer page (English)Read in full. Source for the PDF upload, the 30+ criteria in 6 areas and the two-minute result.
- C Internet Archive copy of the Karvi home page, 10 July 2026Read in full. Source for the July 2026 user counter and the product at that date.
- C Karvi report: State of LinkedIn in Brazil in 2026Read the summary and method sections. Source for the 26,507 analyses from January to May 2026 and the share of Portuguese profiles.
- C Karvi sign-in page (reached from karvi.app/termos, 6 November 2026)Read in full. The address led to the sign-in page. Source for the sign-in options and the outcome claims shown there.
- C Database of failed startups, local exportLocal export. AI-written summaries, not independently verified. Source for the failed predecessors.
Run this company? Send corrections or a verified figure to aiolum@livyden.com