Saddam S.

I design and industrialize complete AI solutions, from the trained model to the mobile application in the hands of the end-user.

My profile is hybrid, and that's precisely what makes it useful: I'm equally comfortable fine-tuning a computer vision model (Segment Anything, Flux), building the GPU backend that serves inference (FastAPI, Modal, Runpod, Docker), processing massive volumes of geospatial data (LIDAR, SwissTopo, 3D reconstruction), as I am integrating it all into a native iOS app (Swift, SwiftUI, UIKit).

Concretely, at Solarsplit, I built a complete automated roof analysis pipeline: processing hundreds of thousands of LIDAR files, 3D reconstruction, calculating surface/slope/orientation, fine-tuning segmentation models, serverless deployment on GPU, and integration into the final product. Before that, 2 years of iOS development in a banking environment (Euro Information) on high-stakes payment apps - which gave me a product rigor that I still maintain today.

What I can deliver for you:

A complete AI pipeline, from dataset to production

An optimized GPU inference service (cost, latency, scalability)

The integration of AI into an existing or new iOS app

Moving from prototype to production for AI/computer vision solutions

Industrializing models that are currently still in the POC stage

I work well when the problem is ambiguous and requires connecting multiple worlds (data, AI, backend, mobile). If you're looking for someone who can deliver a functional end-to-end AI product, let's talk.