Automation discovery
& prioritization.
Where is automation worth the effort?
Understand how work happens. Find the manual effort. Decide which opportunities deserve investment.
Explore the frameworkI led an automation engagement for an insurance company, supervising two colleagues. Over several weeks, we worked with business teams to review approximately 120 processes and build an automation opportunity library.
The insurance assessment illustrates the approach used during the engagement. Supply chain shows how I would apply the same framework in another operating environment. Both interactive assessments use fictional processes and figures.
The engagement / Insurance
My role in discovery and delivery.
- Processes reviewed
- ~120
Approximate scope of the opportunity library
- Colleagues supervised
- 2
Alongside business teams and management
I presented opportunities and their expected benefits to management. After management selected a workflow, I personally built the pilot and provided knowledge transfer and guidance to help the internal team develop further automations.
Lead the discovery
Supervised two colleagues and facilitated workshops to understand how teams worked, including the systems, manual steps and approval handoffs involved.
Support the investment decision
Built the opportunity library and presented use cases and their expected automation benefits. Management selected which workflow to take forward.
Build the pilot
Personally built the solution for the selected opportunity, carrying the work from assessment into delivery.
Transfer the knowledge
Facilitated knowledge sharing and knowledge transfer, with guidance to help the internal team learn how to develop further automations.
A workflow reconstructed from memory
One type of workflow involved transferring spreadsheet information into another system, followed by verification and approval. Looking at the complete sequence helped identify where repeated data entry sat within the wider process.
- 01
Data in a spreadsheet
- 02
Log in and enter data
- 03
Verify the information
- 04
Send for approval
A reconstructed example based on my recollection of the workflows assessed. The original pilot's process name and exact scope are unconfirmed.
01 / The shared approach
Five steps to a considered shortlist.
A repeatable way to turn operational pain points into an investment discussion. Select a step to explore what it establishes.
Where does the team's time actually go?
Work with the people performing the process. Separate hands-on work from waiting, capture demand and exceptions, and understand the pain points before proposing a solution.
Current process map and effort baseline
02 / Apply the framework
Same questions. Different work.
More time for the cases that need judgment.
An illustrative insurance operations team handles policy changes, claims intake and reconciliations. Which activities could release capacity without removing necessary judgment?
Informed by my insurance discovery experience. Processes, figures and recommendations below are fictional examples.
Scroll the table horizontally to compare all columns
| Process / activity | Cases / month | Manual min / case | Potential h / month | Build effort | Initial recommendation |
|---|---|---|---|---|---|
| 8,000 | 6 | 56070% less handling | 5 weeks | Pilot candidate | |
| 4,200 | 12 | 46255% less handling | 8 weeks | Validate first | |
| 20,000 | 1.5 | 40080% less handling | 7 weeks | Pilot candidate | |
| 1,800 | 30 | 18020% less handling | 15 weeks | Redesign first |
Potential capacity = volume × manual minutes per case × assumed handling reduction ÷ 60. Build weeks are illustrative effort estimates, not delivery commitments.
03 / Compare the opportunities
Benefit matters. Readiness changes the decision.
Select a point or a process above to inspect the reasoning. The matrix shows potential handling time released against build effort. Suitability checks sit alongside it.
Indicative build effort
- Pilot candidate
- Validate first
- Redesign first
A promising position is a starting point for discussion. Input quality, controls and exception handling can change the recommendation.
Policy data updates
Rekey approved policy changes across two systems.
potential handling time released
from 800 manual hours / month
Released time is a capacity estimate. Cash savings require a separate plan for how the capacity will be used. Changing this assumption does not resolve suitability checks.
Inside opportunity 01
The process behind the estimate
- 01
Approved request
- 02
Check required fields
- 03
Update policy systems
- 04
Log result / route exceptions
Rules
Defined in exampleApproved changes follow a defined field mapping. Coverage decisions stay with the policy owner.
Inputs
Defined in exampleRequests use a structured form with mandatory fields.
Systems
Confirm before buildConfirm service-account permissions and test both applications' update behaviour.
Exceptions
Defined in exampleMissing data and mismatched records return to a named operations queue.
Assessment recommendation
Pilot candidate
Take a narrow set of approved policy changes into pilot assessment. The rules and inputs are consistent, with a defined route for exceptions.
Evidence needed next
Time a representative sample, confirm access with IT, and test duplicate requests, failed updates and audit logging before approving the build.
Less rekeying may improve consistency between policy systems. Check error and rework rates during the pilot.
04 / The CoE connection
An assessment becomes a delivery decision.
A Transformation CoE can apply the same intake criteria across teams, maintain a shared backlog and make the evidence behind each investment visible.
Prioritized backlog
Opportunity scope, process owner, assumptions and unresolved dependencies.
Pilot approval
Agreed success measures, controls and a bounded scope for testing.
Delivery & adoption
Business and IT owners confirm testing, exception handling and support.
Benefits review
Compare actual handling time, quality and operating cost with the baseline.
The insurance engagement account describes my role in discovery, pilot delivery and knowledge transfer. The interactive assessments use fictional processes and estimates, and do not report the pilot's results or realized savings. The assessment method draws on the 2017 RPA evaluation material.