AI Infrastructure / GPU Engineer

Artificial Intelligence Delivery Unit

Lahore , Pakistan

Apply Now

Posted 3 days ago 29 views Report Job

Job Description: Qualification: Bachelor's or Master's degree (16 years of education) in Computer Science, Computer/Electrical Engineering or a related field from an HEC-recognized university. Preference would be given to foreign qualified candidates with proven track record. Experience: At least 5 years operating GPU/accelerated compute or ML infrastructure at scale, including production model serving. Skills & Competencies: • GPU/accelerator tooling and accelerated runtimes • Kubernetes and GPU scheduling • Model-serving frameworks (Triton, vLLM, TGI or equivalent) • Linux, containers and infrastructure-as-code • Python/Bash Key Performance Indicators (KPIs): • GPU utilisation and cost per training/inference hour • Model-serving latency and availability Job Responsibilities: Key Responsibilities / ToRs: • Provision, configure and optimize GPU compute (cloud and, where justified, on-prem) for training, fine-tuning and inference, maximizing utilization and cost-efficiency. • Build and maintain the model-serving and inference infrastructure (accelerated runtimes, batching, quantization, autoscaling) for production AI workloads. • Manage the AI compute stack - drivers, accelerator libraries, container runtimes and scheduling - and undertake capacity planning across pillar workloads. • Operate the storage and data-movement layer feeding the AI Data Lake in coordination with Data Engineers, without owning data governance (which rests with Wing 3). • Benchmark accelerators and foundation-model hosting options and advise on build-versus-buy for compute. • Document infrastructure standards and contribute reusable platform components. Last date to apply: 12-Nov-2026