Marina Bay at night — a city of signals, gathered and ordered
Workshops · Specialist

Data Harvesting.

Turn the open web into evidence. Build a repeatable architecture that gathers reviews, forums, and competitor data into clean, analysable datasets — with AI doing the heavy lifting.

Two 120-minute live sessionsOnlineHands-onNext cohort starts 28 July
Bring it to your team
← All workshops
01 / Upcoming cohorts

When it runs.

Data Harvesting runs as two live 120-minute sessions, two days apart. Two cohorts coming up.

Booking now

July 2026 Cohort

Session 1 Tuesday 28 July 2026 · 16:00 Session 2 Thursday 30 July 2026 · 16:00
Opens soon

August 2026 Cohort

Session 1 Tuesday 25 August 2026 · 16:00 Session 2 Thursday 27 August 2026 · 16:00
Reserve my place
02 / The why

The evidence is out there. Ready to be gathered.

Your customers are already telling you everything — in reviews, forums, marketplaces, competitor sites, communities. The problem was never the data. It was the days someone spends copy-pasting it into a spreadsheet, and the fact that next quarter you start from scratch. This workshop fixes both. You'll learn to build a real, repeatable harvesting architecture: point it at the right sources, pull clean structured data at scale, and analyse it for what actually matters — so the open web becomes evidence you can use, again and again.

03 / What you'll build

From a question to a repeatable pipeline.

We work end to end — the same path you'll follow every time you run it after.

01

Organise your analysis.

Frame the question first, so the data you gather actually answers it — not a pile of text you hope means something.

02

Select and organise your sources.

Decide where to look and structure it before you pull a thing — reviews, forums, marketplaces, competitor sites.

03

Run and extract a first sample.

A small, fast pull to prove the approach and catch the problems before you scale them.

04

Build a full, repeatable architecture.

The pipeline you can run again next quarter without starting over — the part that turns a one-off into a capability.

05

Run the full extraction.

Gather at scale — clean, structured, deduplicated, ready to analyse rather than tidy up.

06

Content, theme and sentiment analysis.

Turn the raw dataset into patterns you can stand behind — what's said, what it's about, and how people feel.

07

Put your story together.

From a wall of data to a clear point of view someone can act on — the reason you gathered it in the first place.

04 / How it works

Two sessions, two days apart.

A small group, live and online, building a real pipeline together — not watching a demo of one.

Session 1

Frame & Sample

Organise the analysis, choose and structure your sources, and run a first sample so you know the approach works before you scale it.

120 minutes
Session 2

Scale & Analyse

Build the full repeatable architecture, run the extraction at scale, then analyse for content, theme and sentiment — and shape the story.

120 minutes
You leave with the architecture built around a real question of your own — ready to run again.
05 / Who's running it

Carlos Hernandez.

Twenty-five years turning messy real-world signal into evidence businesses can act on, now rebuilt around AI. This is the harvesting craft done properly — not a scraping hack, but a method you can trust and repeat.

More about me
A city seen whole from above at night — a system that knows its world
06 / Reserve a place

Two ways in.

Private cohort

Want it for your team alone, built around your category and your sources? A private session on your own questions.

Let's talk
Online or in person · priced to your team.
Enquire about a private cohort
Stop copy-pasting the web.

Build the pipeline once, run it every quarter.

See all workshops