A practical field guide from Automation Ace.
Short answer: Make Process Intelligence helps you decide what to automate next using data about how work actually happens. Announced at Waves '26 and powered by Celonis process intelligence technology, it combines process mining (how processes flow through your systems) and task mining (how people perform tasks) to reveal bottlenecks, rework, and the highest-value automations. Use it to find candidates, score them on impact and readiness, then build in Make with rules, AI, or agents as the work requires.
Availability, plans, and setup details for Process Intelligence were still emerging at the time of writing; check Make for current access.
Short answer
- Connect the systems and teams involved.
- Map how the process really runs and where it slows down.
- Find repetitive manual steps and rework loops.
- Score candidates on impact, readiness, and risk.
- Build, then measure against the baseline.
What Make Process Intelligence is
Make positions Process Intelligence as the answer to “what should we automate next?”: choose the next automation for its business impact, not because it is the loudest complaint. Make says it maps workflows to uncover bottlenecks, eliminate rework, pinpoint the highest-value automations, diagnose root causes, and show where a person needs to step in before you build. Make is a Celonis company, and the technology comes from Celonis.
Process mining vs task mining
| Technique | What it looks at | What it reveals |
|---|---|---|
| Process mining | Event data from systems (records created, updated, approved, closed) | How processes actually flow, bottlenecks, rework loops, delays |
| Task mining | How people perform tasks on their desktops and in apps | Repetitive manual steps, copy-paste, swivel-chair work |
Why picking the right process matters
- Automating a broken process makes it fail faster.
- The loudest request is not always the most valuable.
- Data shows hidden costs: rework, waiting, and handoffs nobody sees.
- AI amplifies whatever process it runs on, good or bad.
How to find opportunities, step by step
- Pick a business area with a clear outcome, such as order-to-cash or lead-to-close.
- Connect data sources and, where appropriate and agreed with the team, task capture.
- Review the real process map: variants, loops, and where time is spent.
- Find bottlenecks and rework: steps with long waits or frequent returns.
- Find repetitive manual tasks: copying data, checking statuses, re-keying.
- Diagnose root causes before automating: missing data, unclear rules, wrong handoffs.
- List candidates and score them.
Be transparent with teams about task mining: what is captured, why, and how it is used.
Score and prioritize opportunities
| Criterion | Question | Score 1-5 |
|---|---|---|
| Volume | How often does it happen each week? | |
| Time | How much people time does each instance take? | |
| Errors and rework | How often does it go wrong or loop back? | |
| Delay impact | Does waiting here slow revenue or customers? | |
| Rule clarity | Can the steps be written as rules (or is judgment needed)? | |
| Data readiness | Are the systems connectable and the data clean? | |
| Risk | What happens if the automation is wrong? |
Plot impact against effort and start with high-impact, lower-effort items; see the impact vs effort matrix and the automation readiness assessment.
Rules, AI, or agents?
- Clear rules, structured data: a scenario.
- One judgment on unstructured input: a scenario with an AI module.
- Varied, multi-step judgment: an AI agent with tools and approvals.
See Make AI agents vs scenarios and combining agents with rule-based automation.
If you do not have Process Intelligence yet
You can still find opportunities with lighter methods: a discovery workshop, a systems inventory, and an exception list. See process mapping workshops, systems and data inventory, and discovery call questions.
Measure the result
Record a baseline before building (cycle time, manual hours, error and rework rates), then measure again after launch. Process data makes the before-and-after comparison concrete; see defining success metrics before you automate.
Frequently asked questions
What is Make Process Intelligence?
Make Process Intelligence, announced at Waves ’26 and powered by Celonis technology, combines process mining and task mining to show how work actually happens, uncover bottlenecks and rework, and identify the highest-value automations to build next.
What is the difference between process mining and task mining?
Process mining analyzes event data from systems to show how processes flow, where they slow down, and where rework happens. Task mining looks at how people perform tasks in their apps to find repetitive manual steps.
How do I find automation opportunities with Make Process Intelligence?
Pick a business area, connect data sources and agreed task capture, review the real process map, find bottlenecks, rework, and repetitive manual tasks, diagnose root causes, then list and score candidates on impact, readiness, and risk.
How should I prioritize automation opportunities?
Score candidates on volume, time per instance, errors and rework, delay impact, rule clarity, data readiness, and risk, then start with high-impact, lower-effort items.
Can I find automation opportunities without process mining?
Yes. Discovery workshops, systems and data inventories, exception lists, and discovery calls can identify strong candidates, though with less precise data than process mining.
Disclaimer: Make features, module names, plans, and credit usage change over time, and several capabilities announced at Waves '26 (such as MakeOS and Process Intelligence) were in beta or listed as coming soon at the time of writing; confirm current details in the Make Help Center. This article may include links to apps, products, or services; some links may be affiliate links, which means Automation Ace may earn a commission at no extra cost to you.