Independent builder · David S

Real search data for people who automate research, build with n8n, run Apify Actors and work with AI agents.

Guides, n8n workflows and free tools built from 26 Apify Actors I run — with the numbers, the costs and the mistakes left in.

data tools & automations by David S
Research pipeline · recorded run
You / the question

“Is anyone actually searching for Apify MCP?”

Replay of a real run · 2026-0820 s live · 6.8 s cached
Complete · replay anytime
FIG. 01 · live

What a template does while you sleep

example
Complete · replay anytime
  1. Schedule trigger (weekly).
  2. Google Sheets: read the keyword list.
  3. Apify node: run an Actor, wait for the dataset.
  4. Google Sheets: append one row per keyword.
  5. Slack: post a digest with rising and dropped keywords.

Guides by topic

All guides →

What this site is

I build and maintain public data tools on the Apify platform: keyword volumes, competitor research, AI search visibility, YouTube and Amazon research. This site is where I write down how those pipelines actually work — real runs, real costs, real failure modes — and publish the n8n workflows and browser tools I use myself.

Everything here follows three rules:

  • Measured, not copied. Prices, limits and error rates come from my own runs, with the date of the test.
  • Free tools stay free. Tools on this site run in your browser or on static data, so using them costs nobody anything.
  • One page per question. No tag archives, no thin roundups.

Where to start

  • Guides — every published guide, grouped by topic: Apify, n8n and SEO first.
  • About — who writes this and why the domain is called Pavatar.
  • Now — what I am working on this month.
  • Pavatar protocol archive — the history of the 2006 Personal Avatar specification that lived at this address.

Why “Pavatar”?

From 2006 to 2011 pavatar.com hosted the Personal Avatar (Pavatar) specification, a small open protocol for self-hosted avatars. Hundreds of blogs and software projects still link to it. The archive keeps that history accurate and those links working; the rest of the site is new and unrelated to the protocol's original authors.