Research guides
Step-by-step AI guides tagged Research.
Spot and reduce hallucinations in your workflow
Fabrication is not random. It clusters in predictable places, which means you can check the high-risk ou...
Read guideAnalyse a spreadsheet with AI and check its work
AI data analysis is genuinely useful and silently wrong often enough that an unchecked number should nev...
Read guideFind keyword gaps with AI without inventing search volume
Ask an AI for search volumes and it will give you confident numbers that are entirely fabricated. Used f...
Read guideBuild a content brief with AI that a writer can use
AI is genuinely good at the research half of a brief and bad at the judgement half. Knowing which half y...
Read guideBuild a RAG pipeline over your own documents
The demo takes an afternoon. Making retrieval return the right passage on the questions people actually...
Read guideClone a voice responsibly and legally
Voice cloning has real uses: accessibility, localisation, and your own voice at scale. All of them depen...
Read guideRun an AI pilot that produces a decision, not a demo
A pilot that cannot fail is not a pilot. Setting the success threshold before you start is what turns an...
Read guideWrite an AI usage policy your team will actually follow
A restrictive AI policy does not stop people using AI. It stops them telling you, which is considerably...
Read guideWrite SQL with AI against a schema it has never seen
A model will write syntactically perfect SQL against a schema it is guessing at. The query runs, returns...
Read guideEstimate what an AI feature will cost you per month
Per-token pricing looks trivially cheap and routinely surprises people at the invoice. The gap is almost...
Read guideKnow when an agent is the wrong answer
A surprising share of agent projects would work better, cheaper and more reliably as a script with one m...
Read guideResearch with AI without inheriting its mistakes
AI is genuinely fast at finding leads and genuinely unreliable at facts. The workflow that works keeps t...
Read guideWritten to be followed, not skimmed
Every guide here is a procedure someone actually ran, with the steps, the time it takes and the places it goes wrong.
Numbered steps, not prose
Each guide breaks into steps with their own anchor, so you can send a colleague straight to step four. Those anchors are minted once and never move.
Levels you can read without colour
Beginner to Expert, shown as filled pips as well as a label, so the level survives a screenshot, a colourblind reader and a printout.
An honest time estimate
The minutes on each card are the sum of its steps, not a reading speed. Guides without an estimate sort last rather than first — an unknown length is not a short one.
Freshness that means something
“Updated” moves only when a person changed the content — never when someone read it. Sorting evergreen guides any other way ranks whatever was most recently opened.
Filters that admit when they are empty
Every count is measured with your other filters applied, so the number beside an option is what you will actually get. Options that cannot narrow anything are not offered.
Prerequisites up front
What you need before you start, and what you will be able to do at the end, are stated before the first step — not discovered halfway through.
