Basel Saleh

I’m an applied AI / LLM engineer and software engineer with 5+ years across the automotive software stack—from ADAS data tooling and radar programs to production connected-vehicle platforms.

At Irdeto, I initiated and shipped an internal LLM assistant using Retrieval-Augmented Generation (RAG) over technician troubleshooting documentation on AWS (SageMaker/Bedrock), owning design through production rollout to deflect support tickets. I also built an AI-driven diagnostic system for a nationwide hardware recall, cutting diagnostic time by 60% across 500+ faulty units.

I’ve delivered and scaled backend features for a phone-as-a-key platform serving 153,000+ drivers, covering OCR, SSO via AWS Cognito, and firmware upgrades—resolving real field edge cases. I also took a distributed inter-process communication (IPC) system from 0-to-1 and productionized it across devices.

On the edge, I ran an inference lab focused on model quantization and compression for memory-constrained hardware, including llama.cpp self-hosting and KV-cache tuning. Earlier, at Bosch and Continental, I built data pipelines and simulation tooling in Azure, authored ADAS requirements, and supported robust test automation and Trace32 communication.