Omar T.-Senior Software Engineer & Technical Consultant — C++ · Real-Time Systems · ADAS & Robotics

Key Client: MILLA Group – ISFM

• Design and maintain C++ components in a production robotics/autonomous-driving platform (MILLA Group): real-time multi-sensor data-processing pipelines fusing LiDAR, GNSS/IMU and odometry under strict latency and determinism constraints — improving reliability in degraded-signal scenarios by 40%, verified with a custom Python benchmarking and regression suite.
• Root-caused a months-long series of field incidents across three sites to a single sensor-integrity fault, debugging across software, OS, network, and hardware boundaries; shipped a layered fix in a versioned release, hardened with 60+ automated regression tests and an incident-replay harness, and added observability to catch recurrence.
• Built reliable interfaces across a ROS2-based stack and validated on real hardware; delivered PTP time-synchronization bring-up over the sensor network (packet-level debugging with tcpdump) and a runtime/CPU profiling toolkit (perf, Tracy/Perfetto) to meet latency and resource targets on embedded Linux.
• Acted as technical owner: introduced structured root-cause analysis, formal code review and merge-request workflows, incident tracking, and defined technical standards and the validation roadmap through production sign-off.
• Engaged by MILLA Group (autonomous shuttle manufacturer) to integrate and harden a safety-critical AD stack toward production: audited the architecture across perception, HD mapping, and positioning, and delivered a gap analysis with remediation roadmap.
• Design and implement interfaces between perception, localization, and vehicle systems in C++; identify integration risks and drive resolution of cross-subsystem technical issues across the AD stack.
• Standardized the client's software development lifecycle by introducing Agile workflows and CI/CD pipelines, shortening integration and validation cycles.

• Led GNSS/IMU localization quality for a 10 cm-accuracy mapping and 3D data program: designed a three-tier automated assessment framework (signal lock, uncertainty bounds, trajectory plausibility) over large-scale real-world recordings, cutting data-selection time 40% and flagging 15% of recordings as unusable before they entered the map/label pipeline.
• Took C++ perception components from prototype to production on a new embedded Linux target; cut processing time 25% through cache-friendly data structures, multi-threaded parallelization, and profiling (perf/Valgrind) to meet hard real-time budgets.
• Built a vendor-evaluation framework from scratch — hand-labeled golden reference plus automated KPI tooling — that exposed a 20% accuracy gap between data-labeling providers in complex road scenes and directly drove contracting decisions.
• Fronted customer support for a Tier-1 radar-perception platform, resolving over 85% of critical issues within two months; the triage-and-communication approach was adopted as best practice for other customer-facing roles and led to a contract extension.
• Cut new-engineer onboarding from 12 to 6 weeks by creating hands-on perception-stack learning modules, design-rationale sessions, and structured pairing.
• Drove system integration of perception components with motion planning, decision control, basic software, and validation teams — integrating complex software components across multiple subsystems, supported by SIL testing and simulation.

• Three years on parking systems and low-speed maneuvering functions: developed slot-detection algorithms (C11/C++) across multiple generations of Valeo's Parking Assistance System, shipped to SOP on programs for Audi (zFAS), Daimler, BMW, VW, Tesla, JLR, and Renault.
• Owned the full software lifecycle pre- and post-SOP for perception components under the V-cycle, including unit, component, and integration test design.
• Analyzed safety-critical field issues down to their architectural root cause in the detection/motion-control interaction, advocating and delivering robust fixes over surface patches on live OEM programs.
• Built simulation, testing, and visualization frameworks in Python and C++/Qt supporting vehicle-level validation of parking functions; collaborated peer-to-peer with counterpart teams in Stuttgart, Germany.