Anas Mhana

Senior Software Engineer

Munich, Germany anasm87@gmail.com GitHub LinkedIn

Professional Summary

Senior Software Engineer with 11 years of experience at the intersection of high-performance C++ development and Computer Vision. Proven track record in designing and optimizing real-time embedded software for ADAS and communication systems at industry leaders like Bosch, DENSO, and Rohde & Schwarz.

Engineering Projects

Agentic AI Driving Decision Engine

An autonomous vehicle decision-making engine powered by LangChain and Google's Gemini LLMs. This engine uses a tool-calling architecture to gather real-time simulated telemetry and perception data, reason about the driving environment, and output rigid, structured control commands.

Python 3.10LangChainGoogle GeminiPydanticpytest

GitHub

Extended Kalman Filter (EKF) for Sensor Fusion

Unlike standard textbook implementations, this EKF is engineered for production-grade numerical stability, specifically designed to prevent covariance divergence in high-dimensional state spaces and floating-point edge cases.

C++17Eigen3GTestCMake

GitHub

Visual SLAM Front-End (C++)

A robust, mathematically rigorous Visual Odometry (VO) pipeline written in modern C++. This repository represents the perception front-end of a Simultaneous Localization and Mapping (vSLAM) system, engineered to track camera ego-motion across sequential video frames in real-time.

C++17OpenCVEigen3ORBRANSACvcpkg

GitHub

Multi-Threaded Vision Pipeline

3.83x speedup

This project implements a Thread Pool architecture to perform parallel Sobel edge detection. It is designed to maximize CPU throughput while minimizing lock contention and avoiding false sharing at thread boundaries.

C++17Thread PoolSobelCMake

GitHub

Agentic VQA System

This project implements a Production-Grade Visual Question Answering (VQA) Pipeline. The pipeline is designed to take an image (e.g., a dashcam frame) and a text prompt, process them using an open-source Vision-Language Model (VLM), and return a descriptive answer.

PythonPyTorchHugging Face TransformersBLIP

GitHub

Experience

Software Development Engineer

2023 — 2025

Rohde & Schwarz GmbH (via ALTEN GmbH) • München, Germany

Maintained and enhanced high-performance C++ applications for mission-critical radio communication testers.
Implemented robust software design patterns to improve code maintainability and long-term scalability.
Improved software reliability by identifying and resolving complex bugs in the existing codebase.
Optimized core algorithms, significantly reducing processing time and enhancing system performance.

Real-Time Embedded Software Engineer

2022 — 2023

Bosch GmbH (via ALTEN GmbH) • Remote (Lindau, Germany)

Engineered and developed C++17 real-time embedded software components for multi-container, multi-layer appliance systems.
Led the redesign of the appliance power management logic and the pairing mechanism with cloud services.
Ensured system robustness through rigorous bug-fixing, endurance testing, and comprehensive documentation.

Computer Vision Development Engineer

2019 — 2021

DENSO ADAS Engineering Services GmbH • Lindau, Germany

Developed and trained deep learning models for depth completion and 3D reconstruction tasks.
Researched and applied novel computer vision technologies to estimate vehicle depth perception.
Developed real-time visual odometry algorithms and corresponding unit tests in C++.
Validated algorithm performance by defining evaluation metrics and experimenting with real-world data.

Computer Vision Development Engineer

2016 — 2019

ADASENS Automotive GmbH • Lindau, Germany

Developed robust C++ software features and unit tests for an offline mono-camera calibration system.
Designed and built a validation tool to ensure calibration accuracy and reliability.
Created an optical self-diagnosis system by modeling and training traditional machine learning models.

Education

Master of Science in Vision & Robotics (VIBOT)

2014 — 2016

Université de Bourgogne • France

Master’s Thesis: Visual based localization in large-scale environments. Developed a visual localization system using HOG features and linear SVM classifiers.

Master of Robotics Programming Engineering

2011 — 2012

Damascus University • Syria

Computers and Automation Engineering

2005 — 2010

Damascus University • Syria

Video Demonstrations

Visual Based Localization in Large-Scale Environments

Master’s Thesis project developing a robust visual localization system. Engineered a feature extraction pipeline utilizing Histograms of Oriented Gradients (HOG) combined with linear SVM classifiers to enable accurate spatial awareness in large-scale mapping environments.

Autonomous Robotics Navigation (SMART / TurtleBot2)

Developed and deployed Robot Operating System (ROS) scripts for autonomous navigation utilizing SLAM algorithms. Implemented a hierarchical state machine designed to execute complex tasks triggered by real-time object recognition pipelines.

Computer Vision Algorithms Package

Implemented a comprehensive suite of classic computer vision algorithms including feature extraction and image filtering. Engineered a custom graphical user interface using the Qt framework for real-time algorithm visualization and parameter tuning.

MRI-PET Medical Image 3D Registration

Developed a 3D medical image registration pipeline to align MRI and PET scans. Designed algorithms to optimize voxel-based similarity metrics, enabling accurate cross-modality anatomical overlay for advanced diagnostic visualization.

3D Mesh Processing

Engineered 3D mesh processing algorithms for analyzing and manipulating complex geometric structures. Implemented core topological operations to refine and optimize 3D surface reconstructions.