DATA & AI · BEGINNER · NEW
AI Automation Without Code
Turn repetitive office tasks into reliable AI-assisted workflows, with checks that keep a person in control.
Many teams lose hours every week to copy-and-paste work that follows the same rules every time: sorting emails, retyping form answers, summarising notes. This course shows you how to find those tasks, write prompts that return reliable, structured output, and connect AI steps to email, spreadsheets and forms with no-code automation tools. You will build with a human reviewer in the loop, protect sensitive data, and measure whether each automation actually saves time.
- 4modules
- 12lessons
- 3h 41mof material
- Oct 2026last updated
CWClearwork course teamWritten and maintained by the Clearwork data & AI curriculum team
What you will be able to do
- Score your recurring tasks by frequency, time and rules to find the three best automation candidates
- Write five-part prompts that return consistent, structured output a spreadsheet or email step can use
- Check AI output for fabricated or wrong content and set up verification proportionate to the risk
- Build a trigger, filter and action workflow that connects an AI step to email, forms and spreadsheets, with human review before anything goes out
- Measure net time saved and keep automations healthy with named owners, alerts and a monthly review
Who it is for
- Office, operations and admin staff who handle the same emails, forms and spreadsheets every week
- Small-business owners and freelancers who want to automate routine work without hiring a developer
- Team leads who need to judge which AI automations are worth building and how to run them safely
Before you start
- Comfort with email, spreadsheets and online forms; no coding or AI experience needed
- Access to an AI assistant and a free or trial account on a no-code automation platform
- About three hours per week for four weeks
Curriculum
4 modules · 12 lessons · 3h 41m
Module 1Find the Work Worth Automating3 lessons · 47 min
Learn to spot tasks that suit automation, map how they really run today, and pick a first project with a measured baseline.
- Spotting automatable tasks: frequency × time × rulesA simple scoring method that separates tasks worth automating from tasks that only feel tedious.15 min · free preview · worked example · exercise
- Map the process before you automate itWrite down how a task really runs today, including the exceptions, so you automate the real process rather than an idealised one.18 min · worked example · exercise
- Choosing your first automation and measuring the baselinePick a small, visible first project and record how long the task takes today, so you can prove the result later.14 min · worked example · exercise
- Module quiz4 questions with explanations
Module 2Prompts That Produce Usable Output3 lessons · 1h
Write prompts with a clear structure, ask for output that the next step can read, and check results for fabricated or wrong content.
- The five-part prompt: role, context, task, format, examplesA repeatable prompt structure that turns vague requests into consistent results you can run hundreds of times.20 min · worked example · exercise
- Structured outputs that a spreadsheet can useAsk for fixed fields, labels and formats so the next step in your workflow can read the AI’s answer without anyone retyping it.18 min · worked example · exercise
- Checking AI output: hallucinations and verificationWhy AI produces confident errors, which errors matter most for your task, and how to build verification that fits the risk.22 min · worked example · exercise
- Module quiz4 questions with explanations
Module 3Building No-Code Automations3 lessons · 1h
Understand the building blocks every no-code platform shares, connect AI steps to everyday tools, and design review points that keep people in control.
- Triggers, actions and filters: how no-code workflows workThe shared building blocks behind Zapier-, Make- and n8n-style platforms, explained without depending on any one tool’s screens.18 min · worked example · exercise
- Connecting AI steps to email, spreadsheets and formsPractical patterns for the three most common sources and destinations in office automation, including field mapping and typical failure points.22 min · worked example · exercise
- Designing human-in-the-loop reviewWhere to put people in an AI workflow, how to make review fast, and how to decide when review can safely be reduced.20 min · worked example · exercise
- Module quiz4 questions with explanations
Module 4Run Automations Safely and Prove They Work3 lessons · 54 min
Protect sensitive data, measure the real time saved after review and maintenance, and keep workflows healthy as tools and inputs change.
- Data privacy: what not to paste into AI toolsHow to decide what data an AI tool may see, which categories to keep out by default, and how to reduce what you send.20 min · worked example · exercise
- Measuring time saved honestlyCalculate net time saved after review, fixes and maintenance, and track quality so that faster does not mean worse.16 min · worked example · exercise
- Maintaining automations so they keep workingOwners, records, alerts, prompt versions and a monthly review: the routine that stops workflows failing silently.18 min · worked example · exercise
- Module quiz4 questions with explanations
