Case study · Big 4 advisory · PMI
Integration Risk, Sized Pre-Close.
1.2M lines mapped across 19 domains into a defensible CapEx model.
- Lines of code
1.2M
across filtered repositories, mapped in a single automated pass
- Repositories
227
in scope across 43 business and 19 functional domains
- Avg complexity
2.9
function-weighted cyclomatic complexity across 94,180 functions
- Domains
19
functional domains mapped, including undefined code
The challenge
An integration opinion that needed evidence.
The advisory team was asked to assess post-close integration feasibility. Management sessions gave direction but no visibility into how code was distributed across languages and domains.
Without that, re-platforming timelines, translation effort and required skills could only be estimated, not sized.
- 01
How is the codebase actually distributed across languages, stacks and functional domains?
- 02
Where does complexity concentrate, and which components carry the highest rewrite risk?
- 03
Which legacy or specialist-skill dependencies need to be priced into the integration plan?
The solution
The full estate, mapped in one pass.
CodeDD mapped 227 repositories across 43 business and 19 functional domains in a single automated pass. Filtering by stack, language and cyclomatic complexity pinpointed re-platforming candidates and priced the effort, without weeks of engineering interviews.
- 01
Mapped the full codebase
227 repositories scanned in one pass, every language and stack quantified by lines of code.
- 02
Classified functional domains
43 business domains and 19 functional domains assigned across the estate, exposing where integration effort concentrates.
- 03
Graded complexity, function by function
94,180 functions scored by cyclomatic complexity, isolating the minority driving rewrite risk.
- 04
Flagged specialist-skill risk
Legacy languages with a shrinking developer pool identified and priced into the integration plan.
codedd.ai / development activity
Codebase complexity & stack map
227 repositories · 1.2M LoC · 19 domains1.2M
Lines of code
Across filtered repositories
227
Repositories
In scope
2.9
Avg cyclomatic complexity
Function-weighted
19
Domains
Includes undefined
Programming language distribution
- Java70.8% 873.0K LoC
- TypeScript13.1% 161.8K LoC
- JavaScript11.4% 140.3K LoC
- HTML3.1% 38.5K LoC
- Groovy1.1% 13.1K LoC
- CSS0.2% 2.8K LoC
- SQL0.1% 1.7K LoC
Groovy — 13.1K LoC · 1.1% of code · avg CC 3.7 · shrinking developer pool; re-staffing and migration cost factored into integration planning.
Portfolio-wide — domain profile
- Application code46.2% 564.1K LoC
- User interface24.1% 293.7K LoC
- API18.6% 227.3K LoC
- Data processing4.4% 53.4K LoC
- Others (15)6.7% 82.3K LoC
Cyclomatic complexity distribution
94,180 functions across 12,841 files · avg CC 2.9The results
A defensible CapEx model, sized before close.
Exact migration scope sized across 1.2M lines of code and 19 functional domains.
Defensible CapEx model delivered to the client, built on complexity-graded rewrite effort, not estimates.
Specialist talent risk flagged automatically: 13.1K lines of Groovy with a shrinking developer pool, priced into integration planning.
Advisory delivered at deal speed first view in hours, full domain map in days.
“We walked into negotiation with a sized CapEx line, not a guess. That changed the conversation.”
Partner, PMI & Technology Integration, Big 4 advisory firm