Haytham Ichahbane

I work on the Oracle APEX platform at Oracle. Day to day that means building features for the product itself, contributing to strategic enterprise applications that replace external software, and digging into the database when something is slower than it has any right to be.

Most of my professional energy goes into Oracle internals — execution plans, wait events, query rewrite, the boring craft of making SQL fast. I also built a CI/CD pipeline for APEX applications using SQLcl so we can ship database releases the same way the rest of the world ships code: version-controlled, reviewed, automated.

How I think about this work

Databases are interesting because they're where every abstraction eventually has to make peace with physics — disk, memory, locks, ordering. The closer I get to that boundary, the more I learn. The best systems I've worked on were the ones whose authors took performance seriously from the first design doc, not the ones that got "optimized" later.

On the side

I'm writing a multi-model database engine from scratch in C++: three storage backends (row-oriented for OLTP, column-oriented for OLAP, and an LSM tree for key-value and time-series) sharing a single query and transaction layer. The point is less the artifact than the practice — to understand storage trade-offs by living with them.

Working together

If you are building something around Oracle APEX, PL/SQL, Python, C++, or database performance, and want a second pair of eyes — performance audits, CI/CD with SQLcl, architecture review, or just a conversation — say hello. Email is the fastest way.

Also happy to talk to people building serious database infrastructure, even when there is nothing concrete on the table.