NotLinear is a Barcelona-based AI education company that teaches non-technical operators, marketers, and founders how to use Claude in their everyday work. Founder Pietro Montaldo runs it as a one-person business: live workshops, cohort courses on Maven, a weekly Substack newsletter, and custom training for companies.
Everything starts with free live sessions. Pietro hosts two to four a month, and a share of the people who attend go on to join a paid course. So the number of people in each free session matters a lot. Using Apify Actors inside Claude Cowork, he now turns other people's LinkedIn posts into a warm lead list, and every event gets around 200 extra signups.
I never did warm outbound for my webinars because it took a day per post. Now Claude and Apify do the preparation, and I spend half an hour sending the messages.
-- Pietro Montaldo, Founder, NotLinear
The challenge: warm leads in public, and no time to reach them #
Before Apify, every NotLinear signup came from Pietro's own content. He'd post on LinkedIn, send the newsletter, and hope the right people saw it.
He knew there was a better source. Every week, someone with a big audience publishes a post about Claude, AI agents, or go-to-market automation, and hundreds of people comment. Each of those commenters has just said, in public, that they care about the exact topic Pietro teaches.
The problem was time. Opening a post with 300 comments, checking each profile, deciding who fits, and writing a message to each one is a full day of work. Pietro runs a one-person company, so he never did it.
The solution: a Claude skill with two connectors #
The whole workflow runs inside Claude Cowork with two connectors: Apify to pull the data, and Apollo to enrich it. Apify is the largest marketplace of tools for AI, so Claude can pick a ready-made Actor for the data and Pietro never has to build a scraper. Both run on free tiers, which is enough for Pietro's volume.
Here's the flow, the same one Pietro shows live in his workshops:
-
Scrape the commenters. Pietro pastes a LinkedIn post URL into Cowork and asks for everyone who commented, with their comment and profile URL. He doesn't pick the scraper himself: Claude searches Apify Store for the best Actor for the job. It chosePost Comments & Engagements Scraper for LinkedIn , a top-rated option that needs no cookies and no login, so nothing runs from his own LinkedIn account.
-
Enrich the list. Apollo adds title, company, location, and email where it finds one, and asks before it spends credits.
-
Score every lead. Claude rates each person from 0 to 5 against NotLinear's ideal customer profile, using a folder that holds the company description, the offer, and the existing customer list. Company pages are filtered out.
-
Draft the messages. For everyone scored 4 or 5, Claude writes a short DM that references their comment and invites them to a free session on the same topic. The ask is small, so saying yes is easy.
-
Save it as a skill. Pietro then asks Claude to turn the whole conversation into a reusable skill. Next time, he hands it a post URL and gets back the spreadsheet, the CSV, and the drafted DMs.
One live run shows how the funnel narrows. A post with 82 comments gave 68 individual commenters, once 11 company pages and the author's own replies were set aside. Apollo matched 62 of them and found 32 emails. Fourteen scored 4 or 5, and those are the only people Pietro contacts.
He sends the messages himself, as LinkedIn connection requests with a short note. At larger volumes the same list could feed an email tool such as Instantly, but he prefers LinkedIn: the volume is lower, and the messages stay personal enough to keep the response rate high.
I tried nothing else before Apify. There was no 'before.' The task was too slow to do manually, so I skipped it. Apify made it cheap enough to exist.
-- Pietro Montaldo, Founder, NotLinear He runs the skill twice a week, once at the start of the week and once right before a webinar, and focuses on large posts with between 100 and 500 comments, where the volume is.
The results #
- Around 200 additional signups for every free live event
- 48% response rate on the outreach messages
- Roughly 10-20% more signups per event than from organic content alone
- An estimated 23 extra course enrollments , at around $600 per course, or roughly $13,800 in revenue that wouldn't exist otherwise
- A 300-comment post now takes a few minutes of review instead of a full day
- $0 spent on tooling , with Apify and Apollo both on free tiers
Pietro doesn't track every signup with UTM links, so the enrollment figure is his best estimate rather than a dashboard number.
What's next #
The workflow is already a Claude skill, and the next step is a scheduled task. Pietro keeps a sheet of posts worth mining. Claude will check it weekly, run the skill on anything new, and email him the list.
The setup isn't tied to LinkedIn, either. Swap the Actor, and the same skill works on Reddit, X, or YouTube. Apify Store has scrapers for all of them.