a
ajwa129

Choudhry S

@ajwa129

AI Platform Engineer, vLLM, GPU Kubernetes, DevOps, Cloud and Security

Pakistan
English, Urdu
About me
I build and run GPU and LLM infrastructure, on-prem and in the cloud. I wrote IaC for GPU cluster orchestration on KAUST Shaheen III, the Middle East's #1 supercomputer (2,800 NVIDIA GH200s). Real results: vLLM latency cut from 3.2s to 800ms, monthly cloud bill dropped from $10K to $4K, LLMs served to 1,000+ concurrent users on bare-metal Kubernetes with zero downtime for 12 months. Stack: vLLM, Triton, GPU Kubernetes (bare-metal and GKE/EKS/AKS), Terraform, Ansible, Ray, Kubeflow, RAG, voice agents. I run a dual RTX Pro 6000 server at home. On-prem GPU work is my daily reality.... Read more

Skills

a
ajwa129
Choudhry S
Offline • 

See my services

AI Implementation & Deployment
I will build ai voice agent, ai receptionist, ai calling agent using vapi retell
From$120 / projectMore details

Portfolio

Work experience

Peregrine_Suite AI

AI Infrastructure Engineer

Peregrine Suite AI • Full-time

Jul 2025 - Present • 1 yr 2 mos

As Chief Technology Officer – AI & Operations at Peregrine AI Ventures, I lead the strategy, design, and deployment of our end-to-end AI solutions and automation infrastructures. My role bridges advanced technology innovation with operational excellence — overseeing everything from AI architecture and platform integration to process automation and intelligent outreach systems. I drive the development of scalable, high-performance AI systems that power client businesses worldwide, ensuring our solutions deliver measurable impact while remaining reliable, secure, and future-ready. Collaborating closely with cross-functional teams, I translate complex business needs into seamless AI-driven operations.