What is Gojiberry AI?
Gojiberry AI is an AI sales agent that finds people showing buying signals and contacts them on LinkedIn and email. Its Y Combinator company page, checked on 20 October 2026, lists it in the Spring 2026 batch, founded in 2025, based in Paris, with a team of seven. The founders are Pierre-Eliott Lallemant (CEO), Romàn Czerny (CMO) and Dylan Teixeira (CTO), who posts on X as Dylan Txa.
Gojiberry's product summary for AI assistants, last updated 11 October 2026, lists the signals it watches, including people who like or comment on competitors' LinkedIn company content, people engaging with posts on topics or from experts you choose, your profile visitors and company page followers, the most active profiles in your target market, job changes, new decision makers, recent funding, hiring, tech stack, company-level website visitors and lookalikes of your best customers. As of 20 October 2026, pricing starts at $99 per user per month with a free trial, and an optional review mode holds leads for approval before anyone is contacted.
Two things Gojiberry does well stand out. It sells itself with its own product, and its founders publish their growth playbook in unusual detail. This teardown takes those posts at their word on tactics, and checks their revenue figures against each other.
How did Gojiberry use warm signals on LinkedIn?
It contacted only people showing intent. In an X post on 24 April 2026, Czerny described the setup behind the company's first $2M in ARR: five LinkedIn accounts, each sending 30 connection requests and 30 DMs a day, aimed only at warm leads showing real intent. Czerny's summary of the principle is that they used their own tool to grow their own tool.
Txa's post of 4 October 2026 names the signals: people engaging with competitors, changing roles, raising money, or posting publicly about the exact problem Gojiberry solves. Txa says that targeting is the gap between a 1 to 2% reply rate and 25 to 40%. We found no published method for those rates, so read them as the founders' own figures.
Email ran alongside LinkedIn. Czerny's April post lists 6,000 cold emails a day and 295,000 in 90 days, sent from 41 domains and 123 inboxes, in plain text with no links or images, in sequences of two or three emails at most. Czerny puts the infrastructure at about $600 a month and the result at 900+ opportunities.
That is a lot of volume for a small team. Before copying any per-account pace, read our guide on LinkedIn automation risk.
How did Gojiberry's founders post every day?
From several accounts at once, mostly with lead magnets. Czerny's April post describes six posts a day across six LinkedIn accounts: lead-magnet content six days a week and a founder story one day a week, with 788,187 impressions in the previous seven days. Czerny says one lead-magnet post added $5K MRR in under 24 hours at no cost, and that the team replied to every comment.
The same post lists Reddit (14.8M+ views in 12 months, from posting three times a week, telling real stories, offering blueprints and not debating critics), YouTube videos aimed at competitor keywords, about 50K website visitors a month from SEO, three paid LinkedIn influencer posts a week at about $500 each, and Facebook ads.
Txa's post of 12 August 2026 gives the order in which those channels were added, by monthly revenue:
| MRR stage (Txa, 12 August 2026) | Channel added |
|---|---|
| $0 to $6k | Cold email and LinkedIn outreach to people showing intent, run with their own early product |
| $6k to $25k | Reddit: educational breakdowns in SaaS subreddits, 10M+ organic views |
| $25k to $75k | LinkedIn content, YouTube, motion-design videos and free blueprints |
| $75k to $150k | X, B2B influencers, sponsored newsletters and a lifetime affiliate program; Gojiberry joined YC around this stage |
| $150k and up | Meta and Google ads, influencer agencies, and the first serious hires |
Txa's point is that none of the stages replaced the one before. Outbound and content kept running while each new channel was added.
Why did Gojiberry give away a blueprint before asking for a call?
So the resource could do the selling. Czerny wrote in April 2026 that the offer was always the same, a valuable blueprint with no pitch. Txa's October post turns it into a rule: lead with value instead of a calendar link, send a blueprint instead of asking for 15 minutes, and let the free trial become the obvious next step.
The founders use the same mechanic on their own accounts. On 5 April 2026, Czerny marked $1.5M ARR by sharing a short document on how they got there. The 24 April post ends by offering a 0 to $1M ARR go-to-market course to people who repost and comment GTM. As of 20 October 2026, Gojiberry's free resources page lists that course and a warm outbound playbook video.
Pricing is set up for the same funnel. Txa says they settled on $99 a month with a free trial so the funnel runs without dragging anyone onto a call, and Czerny reported 5 to 8 demos a day in April, with about 70% closing to the free plan.
Which Gojiberry ARR figures don't line up?
Several of them. All are self-reported, and some cannot be true at the same time as worded. Each is shown as the page gives it, with its date.
| Date | Where | ARR as stated | Time from zero, as stated | Customers, as stated |
|---|---|---|---|---|
| 25 Mar 2026 | Launch YC post by Czerny | $1.4M ARR; $112K MRR, 44% monthly growth | 9 months | 1000+ paying customers |
| 5 Apr 2026 | Czerny on X | $1.5M | Not stated | Not stated |
| 24 Apr 2026 | Czerny on X | $2M | A few months from €0 | Not stated |
| 17 Jun 2026 | Czerny, reply on X | $3M | Not stated | Not stated |
| 12 Aug 2026 | Txa on X | $4.2M | Under a year | Not stated |
| 4 Oct 2026 | Txa on X | Nearly $5M | About 14 months | Not stated |
| No date shown; checked 20 Oct 2026 | YC company page description | $2.5M, growing 30% a month, profitable | 10 months | 2,000+ customers |
| Checked 20 Oct 2026 | gojiberry.ai | Not stated | Not stated | 3,500+ B2B sales teams |
Two things conflict. First, the same YC page gives $1.4M ARR at 9 months in the launch post of 25 March 2026 and $2.5M at 10 months in its undated description. Nine months before 25 March 2026 is late June 2025, and ten months after that is late April 2026, when Czerny posted $2M. Second, Txa's posts imply a later start: about 14 months before 4 October 2026 is early August 2025, and under a year before 12 August 2026 is after 12 August 2025. Ten months after early August 2025 is early June 2026, between Czerny's $2M post of 24 April and $3M reply of 17 June, so the YC description is consistent with Txa's start date but not with the launch post's.
None of this needs bad faith to explain. The YC page shows no date, so its line may have been written between April and June and never updated. Founders may also count month zero from different points: Txa says the first $10k came from a six-slide deck, before any code was written. ARR Club, which tracks founder-reported milestones, lists the five X figures from April to October. None of them is audited, so treat them as evidence of fast growth and avoid using any one as a benchmark.
What can a small team copy from Gojiberry this week?
The method scales down even if the volume doesn't. These steps are our suggestions for a 1 to 5 person team; the numbers are examples.
Pick three signals and ignore the rest
For example: people commenting on a competitor's posts, people who changed role in the last 90 days, and companies that announced funding. Log which signal surfaced each lead so you can compare replies by signal after two weeks. Our guide to buying signals for small teams covers where to find each one.
Write one blueprint before one more pitch
Two pages on how you do the thing your buyer wants: the steps, the template, the numbers you watch. Offer it in your first message; there is an example message below these steps.
Post one lead magnet a week from the founder's account
Gojiberry ran six accounts; one is enough to test the idea. Ask readers to comment a keyword to get the blueprint, then reply to every comment yourself.
Stay on one channel until it works
Use Txa's stage table as an order of operations. Keep outbound running until replies are steady before you add Reddit, content or paid.
Measure your own reply rate before trusting anyone's
Count replies per signal for two weeks. Judge every vendor's headline figure, Gojiberry's included, against that baseline.
If you want this done for you, Siela's AI SDR can track people who engage with your competitors or keywords on LinkedIn and X and scores every lead hot, warm or cold with a one-line reason. If you are comparing tools, see our Gojiberry alternatives.
Frequently asked questions
Is Gojiberry AI a Y Combinator company?
Yes. Gojiberry AI's Y Combinator page, checked on 20 October 2026, lists it in the Spring 2026 batch, founded in 2025 and based in Paris, with a team of seven. Co-founder Dylan Txa wrote on X in August 2026 that Gojiberry joined YC around the stage when monthly revenue grew from $75k to $150k.
How much does Gojiberry AI cost?
As of 20 October 2026, Gojiberry's site lists a Pro plan at $99 a month, described elsewhere on its site as $99 per user per month, with a free trial. Pro includes two AI agents and up to 1,800 prospects contacted a month. A custom plan for sales teams of three or more is priced on a call.
Who founded Gojiberry AI?
Three founders, per Y Combinator's page: Pierre-Eliott Lallemant (CEO), Romàn Czerny (CMO) and Dylan Teixeira (CTO), who posts on X as Dylan Txa. The launch post says Lallemant and Czerny previously co-founded CoCo AI, a Shopify app sold in 2025, and Teixeira co-founded Edusign, an attendance SaaS also sold in 2025.
What buying signals does Gojiberry AI track?
Gojiberry's product summary for AI assistants, last updated 11 October 2026, lists 12 signal types: competitor engagement on LinkedIn company content, topic engagement, engagement with chosen experts, profile visitors and page followers, top active profiles, job changes, new decision makers, recent funding, hiring, tech stack, lookalikes of your best customers and company-level website visitors. Users can also import their own lists.
How many LinkedIn requests did Gojiberry send a day?
Co-founder Romàn Czerny wrote on 24 April 2026 that the team ran five LinkedIn accounts, each sending 30 connection requests and 30 DMs a day to warm leads. That works out to 150 connection requests and 150 DMs a day across the five accounts. The post does not say how long that pace ran.
How much ARR does Gojiberry AI have?
The latest self-reported figure we found is nearly $5M ARR, posted on X by co-founder Dylan Txa on 4 October 2026, about 14 months from zero by Txa's count. Gojiberry's Y Combinator page, checked on 20 October 2026, still says $2.5M in 10 months. None of the figures is audited.
Sources
- Y Combinator: Gojiberry AI company page (checked 20 Oct 2026 (no date shown))
- Launch YC: GojiberryAI, AI GTM brain that detects buying intent (published 25 Mar 2026)
- Romàn Czerny on X: $1.5M ARR (posted 5 Apr 2026 (UTC))
- Romàn Czerny on X: Gojiberry AI just hit $2M ARR (posted 24 Apr 2026, 23:50 UTC)
- Romàn Czerny on X: reply, we are at $3M (posted 17 Jun 2026 (UTC))
- Dylan Txa on X: We scaled to $4.2M ARR in under a year (posted 12 Aug 2026 (UTC))
- Dylan Txa on X: nearly $5M in ARR (posted 4 Oct 2026, 22:43 UTC)
- Gojiberry AI homepage and pricing (checked 20 Oct 2026)
- Gojiberry AI: product summary for AI assistants (checked 20 Oct 2026)
- Gojiberry AI: free resources (checked 20 Oct 2026)
- ARR Club: GojiberryAI milestones (checked 20 Oct 2026)
- Siela: Gojiberry alternatives (checked 20 Oct 2026)
