AI Automation Opportunity Assessment Process
How to pick which processes to automate with AI: map the work, measure volume and time, score value against risk and difficulty, prove it with a small pilot, and scale what works.
4 teams11 steps2 decisionsFree download
Free to download. Edit it in Flow Studio to change the teams and steps, ask AI for changes, or export to PowerPoint, Word or draw.io.
Step by step
The same flow as a numbered list, ready for an SOP, a BRD or a ticket.
| # | Step | Who | Next |
|---|---|---|---|
| 1 | Goal: use AI where it paysStartLeadership · step 2 | Leadership | step 2 |
| 2 | Map candidate processesVolume per month, minutes per item, error rate, systems usedProcess owners · step 3 | Process owners | step 3 |
| 3 | Score value, feasibility and riskDocument1-5 scores with evidence; plot value vs effortAutomation team · step 4 | Automation team | step 4 |
| 4 | Pick the top one or twoLeadership · step 5 | Leadership | step 5 |
| 5 | Check data and system accessAPIs, data quality, security and privacy constraintsIT · step 6 | IT | step 6 |
| 6 | Feasible?DecisionIT · Yes: step 7, No: next idea: step 4 | IT | Yes: step 7, No: next idea: step 4 |
| 7 | Run a 4-8 week pilotBaseline vs pilot: time, cost, errors, satisfactionAutomation team · step 8 | Automation team | step 8 |
| 8 | Hit the target?DecisionLeadership · Yes: step 10, No: step 9 | Leadership | Yes: step 10, No: step 9 |
| 9 | Idea stoppedEndAutomation team · Done | Automation team | Done |
| 10 | Scale and retrain the teamProcess owners · step 11 | Process owners | step 11 |
| 11 | Automation liveEndProcess owners · Done | Process owners | Done |
When to use it
Use it when leadership asks 'where should we use AI?' and you want a short, defensible list instead of scattered experiments. It works for RPA, workflow automation and AI agents alike.
Who's involved
- Leadership3
- Automation team3
- Process owners3
- IT2
Yours works differently?
Describe your own process in a few sentences and Flow Studio draws it, with your teams and your rules.
Describe your processarrow_forwardFor the technical team
The build decisions behind this flow. Each step above also carries its technical detail.
- 01Discovery: process mining on system logs (e.g. Celonis, Power Automate Process Mining) or task timing samples.
- 02Scoring sheet: value (hours × cost, errors, revenue), feasibility (data access, API availability, variation), risk (customer impact, regulation).
- 03Pilot design: baseline metrics, control group, success threshold, rollback plan.
- 04Scale: shared platform (orchestration, model gateway, monitoring) so each new automation is cheaper than the last.
How to get it right
- 1Start from the work, not the technology: high volume, rules mixed with reading or writing, and a clear cost today.
- 2Score each idea on value, feasibility and risk, and show the scores. It keeps the conversation honest.
- 3Pilot one process end to end for 4 to 8 weeks with a baseline, rather than ten demos.
- 4Count the change for people: who does the remaining work, and what they do with the time saved.
Questions
How do you identify processes to automate with AI?
List candidate processes with their volume, time per item, error rate and cost, then score each on business value, feasibility (data, systems, rules) and risk. Pilot the top one or two against a baseline before scaling.
What's the difference between RPA and AI automation?
RPA follows fixed rules on structured screens and data. AI automation can read unstructured text, documents and conversations and handle variation, often combined with RPA or workflow tools for the system steps.
More process flow templates
- AI feature integrationTaking an AI feature from idea to production: pick the use case, test models against real examples, add guardrails, pass a security review, and roll out in stages with monitoring.
- RAG pipelineHow retrieval-augmented generation answers questions from your own documents: documents are split, embedded and indexed, then each question pulls the best passages and the model answers with citations.
- AI agent with tools and MCPHow an agentic AI system works through a task: it plans, calls tools through MCP servers with scoped access, asks a person before risky actions, checks results and loops until the goal is met.
- MCP server integrationConnecting an internal system to AI assistants through a Model Context Protocol server: design the tools, secure them, test them, review them and roll them out with monitoring.