I build backend systems, and I prove they work before I say they do.
Eight of my patches are merged into projects I had no prior connection to: CERN's ROOT, .NET runtime, systemd, Apache Airflow and the VS Code documentation. Getting a patch into any of them means convincing a maintainer who has never heard of you, with evidence.
The system I spend most time on is a pre-accounting and logistics program for a haulage business. People use it every day while I keep developing it — right now I am closing the gaps that only showed up once it met real data. I also ship mobile apps in Flutter and React Native, and I have taken a real-time service from one instance to two and measured exactly what that bought.
Looking for an internship or a junior backend role.
| The claim | What the number says |
|---|---|
| A seat cannot be sold twice | 20 concurrent requests, one seat → 1 sale, 19 × 409, one held seat in the database |
| It scales horizontally | 2,500 connections → p95 258 ms on one instance, 14 ms on two |
| Course capacity holds under load | 15 applications, capacity 5 → old logic enrolled 15, current logic enrolls 5 |
| Route ordering is worth having | 10 real stops in Kadıköy → 44.3 km as entered, 14.6 km reordered |
| The tests actually run | 285 passing across my repositories, on every push |
 |
 |
 |
 |
 |
 |
 |
 |
Cards marked PRIVATE link to repositories that are not open to visitors. Happy to walk through any of them in an interview.
| Coffee Shop Management | Windows Forms till for a small café. Making the ordering logic testable is how I found that two waiters could open two tabs on one table and only one got billed. |
| Sefer Defteri | The driver's side of the accounting system. Expo, on-device SQLite, document expiry reminders. Works with no signal in a truck cab. |
| Minik Masal | Audio story player for small children, with a parent gate. Written twice: Flutter and Expo. |
| Business Directory API | FastAPI service collecting business listings by province. Paginated endpoints, Alembic migrations, spreadsheet export. |
| Redmine Deployment | Dockerised Redmine with PostgreSQL on AWS EC2. Built with @hzflora. |
| Pansuman Simulator | Unity training simulator for wound dressing. In progress. |
Eight patches merged into projects I had no prior connection to. The pattern is the same each time: I sweep for a class of defect, throw away almost everything the sweep returns, and open a pull request only for what I can prove.
| Merged | What it was |
|---|---|
| root#23002 | A TMVA header unreachable since 2015. I traced the commit that orphaned it and checked every symbol it declared was already covered elsewhere. |
| root#23004 | Two aggregate LinkDef headers that lost their only caller when the build went back to two dictionaries per package. |
| systemd#43300 | Seven man page cross-references pointing at the wrong section. CI went red; I pulled the 5.4 MB log, showed the failure was an unrelated ppc64le timeout, and said so. Merged with the job still red. |
| airflow#71179 | Twenty links that 404 for every reader but open fine for every author. The directory is a git symlink blob and GitHub will not traverse one. 372 candidates went in, 20 came out. |
| vscode-docs#10119 | Three setting IDs whose casing does not match what VS Code registers, so they resolve to settings that do not exist. I checked all 771 IDs behind 1,760 macros. |
| dotnet/runtime#131865 | Eight documentation links whose targets exist but whose relative paths resolve nowhere. From thirty-nine candidates. |
| rustc_codegen_gcc#945 | Two broken links. The useful part was working out I was in the wrong repository: it is a subtree, so a fix landed upstream would be overwritten on the next sync. |
| eclipse-score/communication#853 | Four links in the design docs of the BMW/Bosch/Mercedes automotive platform. One image URL was written hhttp://, so a diagram had never rendered. |
One more worth mentioning: root#23036 was a
report, not a patch. Three settings shipped in system.rootrc that nothing in ROOT reads. Whether
to delete them or wire them up was a maintainer's call, so I filed both options. The engineer who
wrote that subsystem replied "Yes, some parameters were not cleaned up", opened the fix half an
hour later, and it merged. No commit under my name, which for that kind of finding is the honest
outcome.
balcihkutsi@gmail.com · LinkedIn
İzmir, Türkiye · open to remote and hybrid