Five ready-to-use workflows. Bookmark this page, or print it for your desk.
PainDozens of new papers a week. Reading every abstract and filing by hand eats a whole afternoon.
What AI doesAI runs a three-stage filter; you only deep-read the few that matter.
Starter
Read the PubMed export.csv. Score each paper against my hypothesis
(high/medium/low) and list the top 10 worth reading, each with one reason.PainEven the shortlist still has to be read, distilled, and connected — your focus runs out fast.
What AI doesAI batch-extracts key points and builds a knowledge map; you keep your energy for judgment.
Starter
For each of these 10 PDFs, extract: research question, method,
main finding, and relation to my hypothesis. Output one comparison table.Guide in preparation
PainWrong test, skipped assumption checks, figures that miss journal specs — and you get desk-rejected.
What AI doesR with Claude Code: pick the method, check assumptions, and produce publication-grade figures in one go.
Starter
My data: two groups, continuous, non-normal. Recommend a test,
list the assumptions to check, and produce a publication-grade figure in R.Guide in preparation
PainStaring at a blank document — the ideas are there, but the first sentence will not come.
What AI doesAI gives the skeleton, you add the take; then a de-AI pass makes it read human.
Starter
Rewrite this paper as an 800-word plain-language piece a teenager can follow.
Give me the outline first; I will confirm before you expand it.Guide in preparation
PainSlides always slip to the night before the deadline — and the result never satisfies you.
What AI doesAI handles layout, images, and formatting; you own the narrative and the delivery.
Starter
From this paper, produce a 10-slide narrative skeleton: one take-home
message per slide, plus a suggested figure for each.Guide in preparation
If this toolkit helps your research, consider supporting future guides on Ko-fi.
☕ Buy me a coffeeThe full methodology lives in the "Researcher's AI Survival Guide" series.