Amjad Saadeh Logo
Hello, I am

Amjad Saadeh

Software Engineer specializing in image processing, video systems, and AI — building things that see and understand.

$ motto→"progress, not perfection"▊
BerlinSennheiserSenior Engineer
currently exploring:AI Agents·vibe coding·edge AI on embedded platforms
> about

I have been writing code since I was 14 — first for fun, then to understand how the world works through data. With a background in computer science from Freie Universität Berlin and nearly a decade of professional experience, I have moved from microscopy image analysis to autonomous driving to video communication systems. What drives me is the moment where scientific ideas become something tangible and useful.

> read the full story →
> projects
01 / 04
FeaturedComputer VisionEmbedded AINDA2023 - 2025

Real-Time Video AI on Qualcomm SoC

// How do you run object detection at full framerate on embedded hardware?

Architected and shipped the full video AI pipeline for an embedded product — from GStreamer pipeline design through custom model quantization down to on-device inference on the Qualcomm Neural Processing SDK. Bridged the gap between ML models and the constraints of a real SoC: memory budget, computer power and latency targets. Includes integration into toolchain (Yocto/CMake), evaluation harness, and a concept for continuous model retraining and data management.

C / C++GStreamerQualcomm NPU SDKtflite / quantizationOpenCVYoctoCMakeCVATPython
▸Prototype showcased internally - real-time inference and image manipulation demonstrated live on hardware with full video pipeline running on target SoC.
02 / 04
MLDeploymentOpen Source
Doorbell Detector
// end-to-end ML for a real annoyance

Sound-based ML system to detect my doorbell - covering the full lifecycle: data collection & management, model training, and self-hosted deployment. A progressing project, not a notebook demo.

PythonAudio MLData MgmtDeployment
github.com →
03 / 04
MLOpsADASNDA
ADAS ML Pipeline @ CARIAD
// production ML for autonomous driving data

End-to-end automated training and evaluation pipeline for a driver-assistance function. Includes a custom time-series data management framework for car trip data, integrated with Azure ML and internal MLOps tooling.

Azure MLPythonPostgreSQLTerraformMLFlow
04 / 04
DSPOpen Source
pyMUSIC
// signal subspace methods made accessible

Clean Python implementation of the MUSIC algorithm (MUltiple SIgnal Classification) for direction-of-arrival estimation. Shows the algorithmic depth beneath the CV/ML work — rooted in classical signal processing theory.

PythonNumPyArray Signal Processing
github.com →
▸More experiments and open-source contributions on github.com/amjadsaadeh
> skills
Computer Vision
image · video · spatial
OpenCVGStreamerObject DetectionSegmentationImage RegistrationCamera CalibrationPhase Contrast ImagingFluorescence MicroscopyVisual Odometry3D Visualization
CV
ML & MLOps
train · version · deploy
TensorFlow / tfliteModel QuantizationMLFlowDVCAzure MLScikit-learnKerasContinual LearningData VersioningCVAT / Labelstudio
ML
Embedded & Edge
silicon · firmware · real-time
C / C++Yocto / BitbakeQualcomm NPUCMake / ConanTexas InstrumentsOn-device InferenceTorch / tflite runtimeLinux KernelReal-time Pipelines
EMB
Infrastructure
cloud · k8s · pipelines
PythonDockerk3s / KubernetesAzure / TerraformPostgreSQLAnsibleGitHub ActionsREST APIsLinux Sysadmin
OPS
> interests
🧠AI & Agents
📷Image Processing
🎯Computer Vision
🏠Smart Home
🤖Embedded AI
> experience
2023 →
Sennheiser SE & Co. KG
Senior Software Engineer - Video Processing / Embedded
C++OpenCVEmbeddedVideoQualcomm Neural Processing SDK
2020 - 23
CARIAD SE (Volkswagen Group)
Data Scientist / ML Engineer - Autonomous Driving
PythonMLFlowTFLiteAzureML
2019 - 20
Carmeq GmbH
ML Engineer - Driver Assistance Systems
PythonOpenCVSimulink
2017 - 19
Leica Microsystems
Software Engineer - Image Processing
C++PythonDirectXDeep Learning
> full CV →
> contact

Interested in working together or just want to say hello?

Always happy to chat about image processing, ML, or embedded systems.