# Aafaq Rana

> Senior DevOps Engineer and Site Reliability Engineer (SRE), Lead DevOps at Netsol Technologies, Toptal top 3%. 6+ years running production on AWS, Azure, GCP and OCI. Also an AI trainer and AI evaluator (RLHF, pass@k, agent benchmarks). Based in Lahore, Pakistan, open to remote work and relocation (KSA, UAE, EU).

Contact: m.aafaq.rana@gmail.com · https://www.linkedin.com/in/aafaq-rana-369176156/ · https://github.com/aafaq01 · https://www.toptal.com/developers/resume/aafaq-rana

## Roles
DevOps Engineer, Site Reliability Engineer, Platform Engineer, Cloud Engineer, MLOps and LLMOps Engineer, AI Infrastructure Engineer, Forward Deployed Engineer, AI Trainer, AI Evaluator.

## Experience
- Netsol Technologies (Dec 2021 to present): Senior DevOps Engineer and DevOps Team Lead. GPU and LLM serving on Amazon EKS with vLLM, Karpenter, KEDA and DCGM alerting; air-gapped K3s on customer bare metal; SOC 2 infrastructure control owner; moved delivery off Jenkins and SVN, cutting build time about 50%; MCP and LangGraph agents for Kubernetes operations; Cloudflare edge security and DigitalOcean Kubernetes migrations.
- Turing (Jun 2025 to Jul 2026): Senior DevOps, AI Trainer and Evaluator, pod lead. LLM evaluation with RLHF, preference ranking, reward modelling and pass@k on infrastructure-as-code; agent benchmark task review against a 14-dimension rubric.
- Vera Platforms (Jul 2026 to Sep 2026): Infrastructure benchmark engineer. Terraform and AWS tasks for AI agent training, graded against live cloud state.
- Devsinc (Jul 2020 to Dec 2021): Associate Software Engineer. FastAPI services to AWS through Azure DevOps and Terraform.

## Projects
- [Industrial LLM stack](https://github.com/aafaq01/industrial-llm-stack): self-hosted LLM and vision platform on open-source tools. vLLM, TensorRT-LLM and llama.cpp benchmarks (AWQ INT4 doubled single-user speed; 1,121 tok/s with tensor parallelism on 2x T4), YOLOv8n on TensorRT via Triton, Haystack RAG on Elasticsearch (recall@5 0.87), SLO burn-rate alerts, ArgoCD, KEDA, Jenkins, Kubernetes and OpenShift.
- [AI meeting notetaker](https://fathom-rebuild-five.vercel.app): Next.js, Neon Postgres, Deepgram and Claude, built in 24 hours.
- Resilient 3-tier platform on AWS: eight AWS CDK stacks, ECS Fargate, Multi-AZ RDS, security-gated pipelines, chaos-tested.

## Skills
AWS, Azure, GCP, OCI, DigitalOcean, Cloudflare, Kubernetes (EKS, AKS, GKE, OKE, K3s, OpenShift), Terraform, Pulumi, AWS CDK, Ansible, GitHub Actions, Azure DevOps, Jenkins, ArgoCD, Prometheus, Grafana, Loki, OpenTelemetry, Datadog, Sentry, Kafka, PostgreSQL, Python, Go, TypeScript, Bash, vLLM, TensorRT-LLM, Triton, LangGraph, RAG, RLHF, SOC 2.

## Education
BS Computer Science, NUST (SEECS). AWS Certified Solutions Architect Associate. Microsoft Azure Fundamentals.

## More
- [Full profile](https://aafaq-rana.space/about/)
- [Resume PDF](https://aafaq-rana.space/Aafaq_Rana_Resume.pdf)