Summary
Pr Heddar is a Software Architect with 10 years' experience specializing in GenAI, LLM infrastructure, and AI security for regulated industries such as finance, healthcare, and genomics. He designs and scales compliant, air-gapped, GPU-optimized production stacks—bridging MLOps, MLSecOps, and bioinformatics—to deliver explainable RAG and multi-agent systems with strict HIPAA/GDPR controls. Comfortable across cloud, on‑prem and hybrid environments, he operationalizes inference at scale using tools like KServe, Triton, MLflow, Argo CD and advanced optimization techniques (INT8/INT4, GPTQ, LoRA). His work combines deep hardware expertise (NVIDIA H100/H200, AMD MI300X, Gaudi2) with practical defenses—prompt-injection mitigation, vector access controls, and auditing—to enable secure LLM deployments. He also applies lessons from large-scale RAG/LLM deployments at firms like Uber and Meta to inform licensing, reproducibility and low-latency design patterns.
10 years of coding experience