Back-End Developer / Codifying
AI integrations and back-end systems
Integrating AI tools into back-end services, building the systems those products run on, and keeping database operations healthy.
- Company
- Codifying
- When
- Oct 2023 — Present
- Where
- Belgrade
01
Integrating AI tools into back-end services
The work is to bring AI into existing services, not to leave it as a side experiment. Application features call those tools as part of a normal request: input is checked, the service passes the context it already owns, and the result comes back in a shape the rest of the system can store or act on. Timeouts, failures, and the contract between the service and the tool are part of the integration, so the feature behaves like the rest of the back end.
02
Developing robust back-end systems
Alongside the integrations, I build and extend the back-end systems the web applications run on. New behavior gets a clear place to live. Request handling stays predictable from the incoming call to the stored result, and responsibilities stay separated so the next change does not have to cut through everything. The aim is a system other people can keep shipping on.
03
Managing and optimizing database operations
I look after how these systems read and write data. That covers the shape of the data, the operations the application runs against it, and the places where those operations start to slow the product down. Optimization is practical: refine the access path, keep the data layer something the team can reason about, and treat it as part of reliability rather than a later fix.
04
AI features inside the service
Integrations stay inside the back end the product already uses. A feature calls the AI tool through the service, with the same expectations as any other dependency: a clear input, a result the application can store, and a failure path that does not leave the request half-finished.
05
Database work as ongoing care
Database operations are not a one-time setup. As features land, I keep an eye on how data is read and written and adjust the parts that start to cost the application time. The aim is a data layer that stays understandable while the product changes.
AI toolsBack-end systemsDatabases