NimbusIQ - AI Cloud Architect

March 16, 2026 · View on GitHub

AI Dev Days 2026 Hackathon submission | Category: Best Multi-Agent System + Best Enterprise Solution

NimbusIQ is an AI-powered cloud intelligence platform that continuously discovers, scores, and evolves Azure service estates using a ten-agent analysis pipeline powered by Microsoft Agent Framework and Microsoft Foundry. It detects configuration drift, surfaces governance-approved remediation plans, and generates deployable IaC - all requiring human approval before any change is applied.

Dashboard Pane

Architecture

┌──────────────────────────────────────────────────────────────────┐
│  Frontend (React + Fluent UI v9)                                  │
│  Service graph · Recommendations · Approval workflow · Timeline   │
└─────────────────────────┬────────────────────────────────────────┘
                          │ REST / JWT (Entra ID)
┌─────────────────────────▼────────────────────────────────────────┐
│  Control Plane API (.NET 10 / ASP.NET Core)                       │
│  Service groups · Analysis runs · Decisions · RFC 9457 errors     │
└──────────┬──────────────────────────────┬────────────────────────┘
           │ PostgreSQL (EF Core)          │ Agent messages
┌──────────▼──────────────────────────────▼────────────────────────┐
│  Agent Orchestrator (.NET 10 background worker)                   │
│                                                                   │
│  DiscoveryWorkflow ──► MultiAgentOrchestrator (Microsoft MAF)    │
│    Resource Graph        │                                        │
│    Cost Management       ├─ ServiceIntelligenceAgent              │
│    Log Analytics         ├─ BestPracticeEngine (700+ rules)      │
│                          ├─ DriftDetectionAgent                   │
│                          ├─ WellArchitectedAssessmentAgent       │
│                          ├─ FinOpsOptimizerAgent                 │
│                          ├─ CloudNativeMaturityAgent             │
│                          ├─ ArchitectureAgent                    │
│                          ├─ ReliabilityAgent                     │
│                          ├─ SustainabilityAgent                  │
│                          └─ GovernanceNegotiationAgent           │
│                                                                   │
│  IacGenerationWorkflow (Foundry-powered Bicep/Terraform)         │
└──────────────────────────────────────────────────────────────────┘
           All on Azure Container Apps + PostgreSQL Flexible Server
           Managed Identity · OpenTelemetry · Key Vault

Full architecture diagrams: docs/diagrams/

How NimbusIQ Differs from Azure Advisor, PSRules, and Azure Quick Review

Azure Advisor, PSRules for Azure, and Azure Quick Review are discovery and scanning tools - they detect violations and report them. NimbusIQ is an orchestration and decision-support platform that sits above these tools.

Capability	Azure Advisor	PSRule	Azure Quick Review	NimbusIQ
Detect configuration violations	✓	✓	✓	✓
Continuous drift trending	✗	✗	✗	✓
AI-powered reasoning across signals	✗	✗	✗	✓ (6 LLM agents)
Workload-scoped analysis	✗	✗	✗	✓ (Azure Service Groups)
Generate deployable IaC (Bicep/Terraform)	✗	✗	✗	✓
Dual-control approval workflow	✗	✗	✗	✓
Explain WHY issues exist	~Basic	~Pattern-based	~Checklist-based	✓ (AI narrative)
Track value realisation	✗	✗	✗	✓
Auditable agent-to-agent lineage	✗	✗	✗	✓ (A2A tracing)

Key Capabilities

Multi-Agent Analysis Pipeline

Ten specialised agents run in configurable Sequential, Parallel, or Leader protocols. Six agents use Azure AI Foundry GPT-4 for deep analysis; four use deterministic rule-based evaluation. The GovernanceNegotiationAgent mediates cross-agent conflicts before results are committed.

Technology	How it is used
Microsoft Agent Framework (MAF)	`WorkflowBuilder` + `InProcessExecution` drives 10 specialised agents with `UseOpenTelemetry()` instrumentation. BestPracticeEngine (700+ rules) feeds normalised results into the agent pipeline.
Microsoft Foundry	GPT-4 model deployment used by 6 agents for LLM-enhanced analysis and IaC generation (Bicep/Terraform).
Azure MCP	Agent-to-tool protocol for dynamic Azure capability discovery with bearer-token authentication.
Azure Container Apps	Hosts all three services with managed identity, auto-scaling, and zero-downtime revision deployment.
GitHub Copilot	Used throughout development for agent scaffolding, Bicep module authoring, and test writing.

Prerequisites

Azure Developer CLI (azd)
Docker Desktop
.NET 10 SDK
Node.js 20+
Azure subscription with permissions to create resources

Quick Start - Deploy with `azd up`

1. Clone and initialise

git clone https://github.com/lukemurraynz/NimbusIQ.git
cd NimbusIQ

azd init

When prompted:

Environment name: Choose a name (e.g., nimbusiq-dev)
Location: Select your Azure region (e.g., australiaeast)

2. Set required configuration

azd env set NIMBUSIQ_POSTGRES_ADMIN_PASSWORD "YourSecurePassword123!"

3. Provision and deploy

azd up

This single command will:

Create an Azure resource group
Deploy PostgreSQL Flexible Server with a database
Create Azure Container Registry (ACR)
Set up Azure Key Vault for secrets
Create Log Analytics workspace
Deploy Container App Environment
Deploy Azure AI Foundry Hub with GPT-4 models (optional)
Build and push Docker images to ACR
Deploy all three services to Container Apps

4. Access the application

azd env get-values

Open the Frontend URL in your browser to access the NimbusIQ dashboard.

Common Commands

# Deploy code changes without re-provisioning infrastructure
azd deploy

# Deploy a specific service only
azd deploy control-plane-api
azd deploy agent-orchestrator
azd deploy frontend

# View logs from a service
azd logs control-plane-api

# View environment variables and outputs
azd env get-values

# Clean up all Azure resources
azd down

Local Development

# Frontend
cd apps/frontend
npm install
npm run dev

# Control Plane API
cd apps/control-plane-api
dotnet restore
dotnet run --project src/Api/Atlas.ControlPlane.Api.csproj

# Agent Orchestrator
cd apps/agent-orchestrator
dotnet restore
dotnet run --project src/Orchestration/Atlas.AgentOrchestrator.Orchestration.csproj

Infrastructure

The infrastructure is defined using Azure Bicep in infrastructure/bicep/main.bicep and uses Azure Verified Modules where available.

Resources provisioned:

Azure Container Apps - hosting for all three services
Azure Container Registry - private Docker image registry
PostgreSQL Flexible Server - database for service metadata and graph
Azure Key Vault - secrets management
Log Analytics - centralised logging and monitoring
Managed Identities - secure authentication between services
Microsoft Foundry - ML workspace for agent orchestration (optional)
Azure AI Services - GPT-4 and embedding models (optional)
Virtual Network - private networking with private endpoints (optional)

Project Structure

├── apps/
│   ├── control-plane-api/    # ASP.NET Core API (.NET 10)
│   ├── agent-orchestrator/   # MAF agent pipeline (.NET 10)
│   └── frontend/             # React + Fluent UI v9
├── infrastructure/
│   └── bicep/                # Azure Bicep IaC (AVM-based)
├── packages/
│   └── contracts/            # Shared contracts
├── scripts/                  # Deployment and provisioning scripts
├── tests/                    # Unit, integration, contract, e2e tests
├── docs/
│   ├── architecture.md       # Detailed architecture documentation
│   └── diagrams/             # Architecture diagrams (.drawio)
└── azure.yaml                # Azure Developer CLI configuration

NimbusIQ - AI Cloud Architect

Architecture

How NimbusIQ Differs from Azure Advisor, PSRules, and Azure Quick Review

Key Capabilities

Multi-Agent Analysis Pipeline

Drift Detection

AI-Powered IaC Generation

Dual-Control Approval Workflow

A2A Lineage Tracing

Hero Technologies