# Deepak Inugala — Senior SRE · MLOps / LLMOps · AI Infrastructure

> Senior Site Reliability Engineer leading a 6-person SRE team at G42.
> AI Platform Engineering · Agentic AI · GPU Inference Infrastructure · International Delivery.
> 11+ years of infrastructure, Linux, cloud, Kubernetes, and AI platform
> engineering — production LLM serving, MLOps pipelines, agentic AI, and
> sovereign air-gapped LLM Ops.

## Contact

- Name: Deepak Inugala
- Location: Abu Dhabi, UAE
- Email: deepak.1990@hotmail.com
- Phone: +971 50 494 5921
- LinkedIn: https://www.linkedin.com/in/deepak-inugala
- Portfolio: https://deepakinugala.github.io/portfolio/
- Open to: Staff / Principal AI Platform & LLMOps roles, Senior SRE Engineer, Senior DevOps Engineer, Senior Infrastructure Engineer — globally

## Current role

**Group 42 (G42)** — Senior Site Reliability Engineer (03/2023 – Present).
Leads a 6-person SRE team covering GPU platform, LLM/MLOps tooling, and
international client delivery. Sole technical authority for GPU infrastructure
at client sites across Kazakhstan, Angola, and Bahrain; supports Maldives and
Ethiopia remotely. Delivers across GOVINT (classified air-gapped), OSINT, and
Smart Nation programmes.

## Core expertise

- **LLM Serving**: vLLM (PagedAttention, continuous batching, tensor parallelism, prefix
  caching, AWQ/GPTQ), NVIDIA Triton (TensorRT, ONNX, dynamic batching, ensembles),
  LiteLLM (OpenAI-compatible gateway).
- **Agentic AI**: LangGraph (stateful supervisor topology), LangChain, AutoGen, MCP
  (Model Context Protocol) tool servers, n8n workflow automation, OpenClaw, NemoClaw.
- **RAG**: Qdrant, Weaviate, Open WebUI, HuggingFace embeddings, sub-50 ms retrieval.
- **MLOps**: MLflow, Kubeflow Pipelines, Argo Workflows, DVC, Feast, Seldon Core,
  Weights & Biases, Kyverno model admission.
- **GPU / HPC**: NVIDIA H100 / H200 / A100, MIG, DCGM Exporter, KubeRay, CUDA/cuDNN,
  InfiniBand, NVLink, NCCL tuning.
- **Kubernetes & Cloud**: AKS, EKS, RKE2, Rancher, Helm, ArgoCD, Argo Rollouts,
  Kustomize, Azure, AWS, G42 Cloud, Harbor.
- **Air-Gapped LLM Ops**: HF snapshot_download offline, Harbor mirrors, data diode
  transfer, AWQ size reduction, internal MinIO model registry.
- **Observability & Security**: Prometheus, Grafana (30+ dashboards), Loki, Fluent Bit,
  Falco, Trivy, OPA Gatekeeper, Kyverno, Keycloak OIDC, zero-trust NetworkPolicies.
- **IaC & SRE**: Terraform, Ansible, GitLab CI/CD, GitOps, Python (LangGraph/LangChain,
  MLflow SDK, Kubeflow SDK), Bash; incident command, on-call leadership, SLO design.

## Highlighted outcomes

- Deployed LLMs from 7B to 72B parameters; AWQ 4-bit reduced 72B from 144GB to 36GB VRAM.
- Achieved P99 TTFT < 200ms at 1,000+ concurrent requests on vLLM / H100/H200.
- Cut MLOps release cycles from weeks to hours via Kubeflow + Argo + MLflow + ArgoCD.
- Reduced overnight SRE on-call interventions by 40% using LangGraph multi-agent
  automation with human-in-the-loop gates.
- Cut operational toil by 50% via MCP tool servers exposing internal APIs to LLM agents.
- Led FGB–NBAD post-merger IT integration at First Abu Dhabi Bank with zero major
  incidents; 35% reduction in repeat incidents via improved incident response.
- Managed 3,000+ physical Linux servers at HCL (UIDAI programme, Bengaluru) with
  on-prem-to-cloud migration delivered in under 2 hours downtime and 0% data loss.

## Experience timeline

- 03/2023 – Present: **Group 42 (G42)**, Abu Dhabi — Senior Site Reliability Engineer
  (leads 6-person SRE team; GPU platform, LLM/MLOps, international delivery).
- 07/2019 – 03/2023: **Group 42 (G42)**, Abu Dhabi — Site Reliability Engineer
  (built the AI platform foundations — Kubernetes, GPU infrastructure, observability).
- 01/2018 – 07/2019: **First Abu Dhabi Bank (FAB)**, Abu Dhabi — Cloud Engineer
  (FGB–NBAD post-merger IT integration, AWS architectures, Cloudera CDH administration).
- 05/2017 – 01/2018: **HCL Technologies**, Abu Dhabi — Technical Specialist
  (Daman health insurance account, Linux + virtualization + incident response).
- 08/2015 – 04/2017: **HCL Infosystems**, Bengaluru, India — Linux Engineer
  (UIDAI / Aadhaar programme, 3,000+ physical RHEL/CentOS servers, on-prem-to-cloud migration).
- 10/2012 – 06/2013: **Polaris Consulting & Services**, Hyderabad, India —
  Technical Support Engineer (Videocon d2h account, Linux and network operations).

## Education

- MBA, Jawaharlal Nehru Technological University (10/2012 – 04/2015).
- B.Tech, Jawaharlal Nehru Technological University (08/2008 – 06/2012).

## Certifications

- Microsoft Azure Administrator Associate (AZ-104) — valid through 07/2027.
- Microsoft Azure Solutions Architect Expert (AZ-305) — valid through 07/2027.
- AWS Solutions Architect – Associate.
- Certified Kubernetes Administrator (CKA) — renewal in progress.
- G42 Cloud Certified Engineer.
- Red Hat Certified Engineer (RHCE).
- Red Hat Certified System Administrator (RHCSA).
- Cloudera Certified Administrator for Apache Hadoop.
- GitOps Certified Fundamentals (Codefresh / Argo).
- Cloud Architecture: Design Decisions (LinkedIn Learning).

## Resumes (role-tailored PDFs)

- LLMOps / MLOps Engineer: /resumes/LLMOps_MLOps_Engineer.pdf
- Senior SRE: /resumes/Senior_SRE_Engineer.pdf
- Team Lead SRE: /resumes/TeamLead_SRE_Engineer.pdf
- Senior DevOps: /resumes/Senior_DevOps_Engineer.pdf
- Senior DevSecOps: /resumes/Senior_DevSecOps_Engineer.pdf
- Senior Cloud: /resumes/Senior_Cloud_Engineer.pdf
- Senior Kubernetes Administrator: /resumes/Senior_Kubernetes_Administrator.pdf
- Senior HPC Engineer: /resumes/Senior_HPC_Engineer.pdf

## Languages

English & Hindi (Full Professional) · Telugu (Native).