Automation & Workflows — Free AI Prompts for M365 Copilot

Design an approval workflow for [PROCESS]. Include: (1) Form fields, (2) Routing rules (by amount, type, department), (3) Escalation if unapproved within [TIME], (4) Notifications (submitted, approved, rejected, escalated), (5) Audit trail. Power Automate actions specified.

Design an automated workflow that generates a [FREQUENCY — daily / weekly / monthly] [REPORT TYPE] report. Include: (1) Data sources to pull from, (2) Calculations and transformations needed, (3) Report format and layout, (4) Distribution list and delivery method, (5) Schedule and timing, (6) Error notification if report fails, (7) Archive and retention policy.

Calculate the ROI for automating [PROCESS]. Include: (1) Current manual effort (hours per week, people involved, error rate), (2) Automation development cost (hours to build), (3) Ongoing maintenance estimate, (4) Time savings per week after automation, (5) Break-even point, (6) Annual ROI percentage, (7) Non-financial benefits (accuracy, speed, satisfaction). Present as a one-page business case.

Write a standard operating procedure for [AUTOMATED PROCESS]. Include: (1) Purpose and business context, (2) How the automation works (plain English), (3) Normal operation — what to expect, (4) How to monitor it, (5) What to do when it fails (manual fallback steps), (6) How to modify or update it, (7) Owner and review schedule. Write for someone inheriting this automation.

Design a chatbot conversation flow for handling [USE CASE — IT password resets / leave requests / FAQ / order status]. Include: (1) Welcome message and intent detection, (2) Main conversation branches, (3) Questions to ask and in what order, (4) Integration points (API calls, database lookups), (5) Handoff to human agent triggers, (6) Fallback and error messages. Map as a decision tree.

Design a data pipeline from [SOURCE] to [DESTINATION] for [PURPOSE]. Include: (1) Data extraction method, (2) Transformation rules, (3) Loading strategy (full vs incremental), (4) Schedule, (5) Error handling and alerting, (6) Data quality checks, (7) Monitoring dashboard. Consider volume, freshness, and reliability.

Design an automated onboarding workflow triggered when HR adds a new hire. Steps: (1) Create accounts and assign licences, (2) Add to groups and distribution lists, (3) Send welcome email with setup instructions, (4) Create onboarding task list, (5) Schedule orientation meetings, (6) Assign buddy, (7) Day-30 check-in reminder to manager.

Analyse this manual process: [DESCRIBE PROCESS]. Map it for automation: (1) Current steps with time per step, (2) Which steps can be automated vs need human judgment, (3) Recommended automation tool (Power Automate, script, API), (4) Trigger event, (5) Data flow between steps, (6) Exception handling, (7) Estimated time savings. Present as a process flow.

Design monitoring alert rules for [SYSTEM/APPLICATION]. For each alert define: (1) What to monitor (metric/log pattern), (2) Threshold or condition, (3) Severity level (info/warning/critical), (4) Who gets notified, (5) Expected response action, (6) Suppression rules to prevent alert storms. Target: actionable alerts only — every alert should require a response.

Design a Power Automate flow for [USE CASE — approval workflow / email notification / data sync / form processing]. Include: (1) Trigger type, (2) Each action step with configuration, (3) Conditions and branches, (4) Error handling and retry logic, (5) Notification to stakeholders, (6) Testing checklist. Specify which connectors are needed.

Design auto-response rules for [SCENARIO]. Include: triggers, templates by sender type, escalation rules, follow-up sequence, exclusion rules to prevent loops.

Design a notification system for [TOOL/PROCESS] that reduces noise. Include: (1) Notification categories by urgency, (2) Channel routing rules (email for urgent, Teams for FYI, digest for low), (3) Frequency caps, (4) Quiet hours, (5) User preference settings, (6) Escalation for unacknowledged critical alerts.

Design an integration between [SYSTEM A] and [SYSTEM B] for [BUSINESS OUTCOME]. Include: (1) Data fields to sync and direction, (2) Sync frequency (real-time, scheduled, on-demand), (3) Authentication method, (4) Error handling and conflict resolution, (5) Data transformation rules, (6) Testing approach, (7) Monitoring and alerting. Consider security and data privacy.