Skip to content
View arunsanna's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report arunsanna

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
arunsanna/README.md

πŸš€ Arun Sanna

Enterprise AI Architect | Multi-Cloud Platform Engineering | Production LLM Infrastructure

LinkedIn Website GitHub


πŸ‘‹ Hi there

I build and ship enterprise AI systems at production scale β€” LLM infrastructure, RAG platforms, multi-agent systems, and multi-cloud Kubernetes platforms serving millions of users.

Over 12 years across financial services (Citibank, PCI-DSS payment infrastructure), retail (McDonald's, Chick-fil-A), healthcare, and highly regulated enterprise environments. I take complex, high-stakes systems from concept to production fast, with security and compliance built in from day one.


πŸ”¬ Current Focus

  • Production AI: Vision-LLM document processing pipelines (95% faster, 99.5% accuracy), Gemini-on-Vertex AI advisory applications, RAG systems, multi-agent platforms
  • MCP Ecosystem: Building MCP servers (aws-sage-mcp, gemini-image MCP) and agent security tooling (AgentMesh-Bench: 60,000+ adversarial attacks)
  • Multi-Cloud Platforms: Kubernetes at 20,000+ container scale across AWS, Google Cloud, and Azure
  • AI Security Research: Published in peer-reviewed journals (JCSTS) and IEEE CLOUD 2026 on AI agent supply-chain security

πŸ’» Notable Projects

Enterprise Gemini AI Advisory Application

Sole architect and builder of an enterprise Gemini-on-Vertex AI application serving 500+ program teams. Reduced a 30-day cross-functional review process to under 5 minutes. Prototype to production in under 60 days.

Production Vision-LLM Document Processing

Built AI document processing pipeline using vision LLMs β€” 95% timeline reduction, 99.5% accuracy, handling handwriting and rotated documents. MVP to production in under 60 days.

Provider-Fraud Detection (Award-Winning)

91.34% blind provider-fraud detection across 33,000+ providers using behavioral signals and Neo4j evidence graphs. Won 2nd place at a national hackathon.

Enterprise Kubernetes Platform (20K+ Containers)

Creator and lead architect of the largest enterprise Kubernetes platform-as-a-Service β€” 20,000+ containers across 20 clusters, 150+ teams, 700+ applications, 100% uptime, zero breaches. Software Factory pattern with 50+ Terraform modules and 50+ Helm charts.

Switchboard β€” Multi-Agent Platform

Production multi-agent platform orchestrating Claude, Codex, Gemini, and Copilot for native iOS, macOS, and web clients. Single gateway, PTY shells, tasks, meetings, tracing.

Commercial Cloud Modernization (Fortune 500)

Cloud solutions architect for Citibank (PCI-DSS, fraud detection, real-time event streaming), McDonald's, Chick-fil-A, American Heart Association (HIPAA), Ditech Mortgage, and Ellucian. 25% operating cost reduction, 60% fewer production risk alerts.


πŸ› οΈ Skills & Technologies

  • AI/ML: Large Language Models (LLMs), RAG, Vertex AI, Gemini, NVIDIA GPU/CUDA, Multi-Agent Systems, MCP
  • Cloud: AWS, Google Cloud (GKE, Anthos, Vertex AI), Azure
  • Kubernetes: EKS, GKE, AKS, Helm, ArgoCD, Istio
  • Infrastructure as Code: Terraform, OpenTofu
  • AI Security: Agent supply-chain security, adversarial attack evaluation, Zero Trust
  • Languages: Python, Go, TypeScript
  • Data: PostgreSQL, Neo4j, Snowflake, Databricks, BigQuery
  • Observability: Grafana, Prometheus, Dynatrace, Splunk

πŸ“ Publications & Research

Peer-Reviewed:

  • Detecting False Assurance in Cloud Control Inheritance Using a Federated Compliance Integrity Graph - IEEE CLOUD 2026 (Regular Full Paper, Cloud Security & Zero Trust session). Single author. DOI, pp. 473-482.
  • Cross-LLM Generalization of Behavioral Backdoor Detection in AI Agent Supply Chains - Journal of Computer Science and Technology Studies (JCSTS), peer-reviewed. arXiv:2511.19874

Conference Talks:

  • Infrastructure-Layer Attack Surface Reduction for MCP-Based AI Agents - InnoAI 2026 (Thailand). Speaker and session chair.
  • AI-Powered FDA Inspection Data Extraction - 13th Annual FDA Scientific Computing Days 2026.
  • DevSecOps for Mission-Critical Systems - AFCEA Alamo ACE 2023.

Trade Publications:

  • Context Engineering: The Next Frontier in AI-Driven DevOps - DevOps.com (6.5M+ monthly visitors).
  • The Orchestration Layer: Why Your AI Agent Needs a Nervous System - Techstrong.ai.
  • Cross-LLM Security for AI Agent Supply Chains - DevOps.com.
  • AWS Sage: Secure MCP Server for AI Agents - DevOps.com.

5 additional papers/articles in progress for 2026.


πŸ† Certifications

  • Certified Kubernetes Administrator (CKA)
  • AWS Certified Solutions Architect Professional
  • Google Cloud Professional Cloud Architect
  • Azure Solutions Architect Expert
  • NVIDIA Certified AI Infrastructure Specialist
  • Palantir Certified Foundry Expert

πŸ“« Let's Connect


Building AI systems that ship.

Pinned Loading

  1. aws-sage-mcp aws-sage-mcp Public

    The aws-mcp project is a Python-based application designed to interact with AWS services using the Model Context Protocol (MCP).

    Python 5 3

  2. arunlab/k8s-sherlock arunlab/k8s-sherlock Public

    K8s-Sherlock is an open-source Kubernetes pod designed for debugging and diagnostics. As a swiss-army knife for your Kubernetes cluster, it comes pre-loaded with a host of tools to help you diagnos…

    Shell

  3. MicroEKS MicroEKS Public

    MicroEKS is a lightweight, interactive tool for creating local Kubernetes environments that mimic AWS EKS clusters on macOS. It uses MicroK8s running in a Multipass virtual machine to provide a ful…

    Go 1 2

  4. AgentDefense-Bench AgentDefense-Bench Public

    A comprehensive security benchmark for evaluating infrastructure-layer defenses in MCP-based AI agent systems

    Python 14 6