Practical AI guides
Let's discuss your project →Codex Automations: What Changes When AI Runs Recurring Workflows
Practical English guide to Codex Automations: synthetic data, a copy-paste prompt, manual testing, documented scheduling, pause rules, and a control log.
Learn more →Codex Settings: What to Configure Before Assigning Tasks to an AI Agent
A practical lab for configuring Codex before the first task: safe defaults, sandbox and approval policy, a harmless synthetic run, checks, and troubleshooting.
Learn more →AI meeting notes: separate decisions, proposals and actions
A practical guide to AI meeting notes: keep decisions, proposals and actions separate, with a transcript, prompt, review table and error check.
Learn more →Compare quotations with AI: a three-offer exercise
A step-by-step AI workflow for comparing three quotations without inventing data, with simulated offers, a prompt, totals, questions, and checks.
Learn more →Routing Tickets with AI: a Practical Trial with Categories and Abstention
A practical lab for routing support tickets with AI: three internal destinations, abstention rules, five synthetic messages, and a human review register.
Learn more →How to evaluate an AI assistant with a small test set before switching models
Build a repeatable A/B test for two AI assistants with three synthetic documents, six questions, a reusable prompt, and a scored evaluation sheet.
Learn more →Extracting Order Data with AI: Missing Fields, Units, and Checks Before Import
A practical lab for extracting order data with AI: strict schema, null for missing fields, a conversion case, validation checks, and a pre-import register.
Learn more →Duplicate master data: use AI to suggest corrections without merging records by mistake
A practical lab with a six-row synthetic CSV, a proposal-only prompt, a JSON schema, and a separate review register to stop AI from merging records by mistake.
Learn more →Product Pages with AI: How to Write Descriptions Without Inventing Features
A hands-on lab for AI product-page copy: build a verified source sheet, use a strict prompt, and review every claim before publishing.
Learn more →Translating a Technical Site with AI: Keeping the Glossary, Numbers, and Promises Aligned
A practical workshop for translating a technical site with AI while keeping glossary terms, numbers, support hours, limits, and promises aligned.
Learn more →Classifying Support Tickets with AI: Categories, Abstention, and Human Review
A practical guide to AI ticket triage with a small taxonomy, abstention rules, human review, a nine-ticket lab, JSON prompt, and separate coverage/correctness metrics.
Learn more →Internal documentation AI assistant: how to verify sources, limits, and a useful “not found”
Practical lab for testing an AI assistant on internal docs: source-only prompt, S1/S2 cases, explicit limits vs missing data, record template, and checks.
Learn more →