Test AI on one task you already do.
Choose the task, decide what information AI may receive, and measure time, corrections, and quality for 30 days.
Your plan includes
- One repeated task and the part AI will assist with
- Rules for data use and the person who checks the result
- A result to measure over the next 30 days
AI Work Lab
Choose one repeated task to test with AI.
Select a role and task below. Your choices and 30-day progress stay on this device.
30-day experiment
Measure the result before using AI on more work.
0 of 4 weeks completed
Week 1
Set the baseline
Complete turn discovery notes into a client brief once using your current process. Record minutes from source material to reviewed output.
Output: A real before-measurement and one safe sample
Week 2
Design the review loop
Give AI only the approved input. Ask it to group evidence into goals, constraints, risks, unknowns, and proposed next steps. Keep the human checkpoint explicit.
Output: A reusable instruction and review checklist
Week 3
Run repeated tests
Use the workflow 3 times this week. Record time, corrections, and where judgment changed the draft.
Output: A short evidence log, not a best-case demo
Week 4
Keep, change, or stop
Compare the result with your baseline. Keep the workflow only if the selected outcome improves without weakening review or data handling.
Output: A defensible SOP and one verified career story
Evidence and ROI checkpoint
Did the workflow improve real work?
Compare completed work with the baseline. Count human review time and failed outputs, not only the fastest example.
Current baseline
30 min/output
Includes drafting, checking, and rework.
Show this project to an employer.
Record the original task, what AI assisted with, what you checked, the measured result, and what you would change next.
Role guides
See the review rules and evidence expected in your field.
Each guide compares three tasks and names the information, decision, and professional responsibility that stay under human control.
AI for Accountants
Start with a reviewed finance workflow, not a financial conclusion.
Open role guideAI for Consultants
Use AI to organize client evidence, not to invent the client story.
Open role guideAI for Managers
Let AI organize the record; keep the management decision human.
Open role guideAI for HR professionals
Use AI to improve the process, not to decide who deserves the job.
Open role guideAI for Software engineers
Start where the output can be tested before it reaches production.
Open role guideAI use by Canadian workers
35.9% used generative AI at work, but use alone does not prove better work.
A useful example shows the original task, what AI changed, what a person checked, and whether time or quality improved.
- 35.9%
- of Canadian workers used generative AI at work in the previous 12 months in March 2026.
- 65.6%
- usage in professional, scientific, and technical services, the leading knowledge-work sector.
- 19.2%
- of Canadian businesses used AI to produce goods or deliver services in the 2026 survey.
Responsible use
Decide what data AI may receive before writing the prompt.
Purpose
Use AI for a defined work purpose, not because a tool is available.
Minimum data
Use the least sensitive information required to test the workflow.
Human ownership
Keep the accountable professional responsible for facts, decisions, and final delivery.
Evidence
Measure quality and review effort, not only how quickly a first draft appears.
CoachGPT provides planning and educational support, not legal, privacy, employment, accounting, or other regulated professional advice.
Questions people ask
Using AI as a professional in Canada
What does AI literacy mean for a Canadian professional?+
It means being able to choose an appropriate work task, protect data, verify output, retain professional judgment, and measure whether the workflow improves a real result. It is broader than knowing how to write prompts.
Can I use confidential client or employee information in an AI tool?+
Do not place confidential, personal, privileged, or regulated information into a consumer AI tool. Follow your employer's approved-tool policy, contracts, professional obligations, and applicable Canadian privacy requirements.
Will this assessment tell me which AI product to buy?+
No. It helps you define a useful workflow and its data boundary before choosing a tool. Product selection should follow the task, security requirements, review process, and evidence you intend to collect.
How do I show AI skills to a Canadian employer?+
Show a verified example: the original business task, what AI assisted with, what you personally reviewed or changed, the measured result, and the safeguards you used. Do not present generated output as independent expertise.
Want to use this project in a job search?
Turn the measured result into a work sample or interview example.