Skip to main content

Play Catalog

The complete catalog of FrootAI Solution Plays. Each play is a production-ready AI blueprint with DevKit (agents, skills, instructions), TuneKit (config), infrastructure (Bicep), and evaluation pipelines. See the Solution Plays Overview for how plays are structured.

Full Play Table​

#NameCategoryComplexityKey Azure Services
01Enterprise RAGRAGIntermediateAzure OpenAI, AI Search, Blob Storage, App Service
02AI Landing ZoneInfrastructureIntermediateVNet, Private Endpoints, Key Vault, Managed Identity
03Deterministic AgentAgentsIntermediateAzure OpenAI (temp=0, seed), Structured Output
04Call Center Voice AIVoiceAdvancedSpeech Services, Azure OpenAI, Event Hub, Container Apps
05IT Ticket ResolutionAgentsIntermediateAzure OpenAI, ServiceNow connector, Logic Apps
06Document IntelligenceDocumentIntermediateDocument Intelligence, Azure OpenAI, Blob Storage
07Multi-Agent ServiceAgentsAdvancedAzure OpenAI, Semantic Kernel, Container Apps
08Copilot Studio BotCopilotBeginnerCopilot Studio, Dataverse, Power Platform
09AI Search PortalRAGIntermediateAI Search, Azure OpenAI, App Service, CDN
10Content ModerationSecurityIntermediateContent Safety, Azure OpenAI, Event Grid
11AI Landing Zone (Advanced)InfrastructureAdvancedHub-Spoke VNet, Firewall, DNS, Policy, Governance
12Model Serving on AKSInfrastructureAdvancedAKS (GPU nodes), vLLM, KEDA, Container Registry
13Fine-Tuning WorkflowFine-TuningAdvancedAzure AI Foundry, Azure OpenAI, MLflow
14Cost-Optimized AI GatewayInfrastructureIntermediateAPIM, Azure OpenAI (multi-region), Redis Cache
15Document Processing (Vision)DocumentIntermediateGPT-4o Vision, Document Intelligence, Blob Storage
16Copilot Teams ExtensionCopilotIntermediateTeams, Bot Framework, Azure OpenAI
17AI ObservabilityObservabilityIntermediateApplication Insights, Log Analytics, KQL, Dashboards
18Prompt OptimizationAgentsIntermediateAzure OpenAI, DSPy, A/B Testing
19Edge AIEdgeAdvancedONNX Runtime, IoT Hub, Container Instances
20Real-Time AnalyticsReal-TimeAdvancedEvent Hub, Stream Analytics, Azure OpenAI, Cosmos DB
21Agentic RAGRAGAdvancedAzure OpenAI, AI Search, Semantic Kernel, Function Calling
22Swarm OrchestrationAgentsAdvancedAzure OpenAI, AutoGen/CrewAI, Container Apps
23Browser AgentAgentsAdvancedAzure OpenAI, Playwright, Container Apps

Category Breakdown​

🔍 RAG & Search — Plays 01, 09, 21​

Ground LLM responses in your data using retrieval-augmented generation:

  • 01 — Enterprise RAG: The foundational RAG pattern — chunking, embedding, hybrid search, semantic ranker, citations. Start here.
  • 09 — AI Search Portal: A user-facing search experience with faceted navigation, filters, and AI-powered answers.
  • 21 — Agentic RAG: RAG with agent autonomy — the agent decides when to search, what to search, and how to combine results.

🤖 Agents — Plays 03, 05, 07, 18, 22, 23​

From single deterministic agents to distributed swarms:

  • 03 — Deterministic Agent: Zero-temperature, seed pinning, structured output. When you need consistent, verifiable results.
  • 05 — IT Ticket Resolution: Classify, route, and resolve IT tickets with AI. Integrates with ServiceNow.
  • 07 — Multi-Agent Service: Supervisor routes to specialist agents. The foundation for complex agentic workflows.
  • 18 — Prompt Optimization: Systematically improve prompts using DSPy optimizers and A/B testing.
  • 22 — Swarm Orchestration: Distributed agent teams with topology management and conflict resolution.
  • 23 — Browser Agent: AI agents that navigate and interact with web pages using Playwright.

🏗️ Infrastructure — Plays 02, 11, 12, 14​

Production-grade Azure foundations for AI workloads:

  • 02 — AI Landing Zone: Secure foundation — VNet, private endpoints, managed identity, RBAC.
  • 11 — AI Landing Zone (Advanced): Hub-spoke networking, Azure Firewall, DNS, governance policies.
  • 12 — Model Serving on AKS: Self-host open-source models on GPU-enabled AKS with vLLM.
  • 14 — Cost-Optimized AI Gateway: APIM-based gateway with semantic caching, model routing, token metering.

🔒 Security — Play 10​

  • 10 — Content Moderation: End-to-end content safety pipeline with Azure AI Content Safety.

🎙️ Voice — Play 04​

  • 04 — Call Center Voice AI: Real-time STT → LLM → TTS streaming for voice-based AI assistants.

📄 Document Processing — Plays 06, 15​

  • 06 — Document Intelligence: Extract structured data from documents using Azure Document Intelligence.
  • 15 — Document Processing (Vision): Multi-modal document analysis with GPT-4o Vision capabilities.

🧩 Copilot Extensions — Plays 08, 16​

  • 08 — Copilot Studio Bot: Build a Copilot Studio bot with custom topics, generative answers, and Dataverse.
  • 16 — Copilot Teams Extension: Extend Microsoft Teams with an AI-powered bot using Bot Framework.

📊 Observability — Play 17​

  • 17 — AI Observability: Monitor AI systems with custom KQL queries, dashboards, and alerting.

🧪 Fine-Tuning — Play 13​

  • 13 — Fine-Tuning Workflow: End-to-end LoRA/QLoRA fine-tuning with data prep, training, and evaluation.

📡 Edge & Real-Time — Plays 19, 20​

  • 19 — Edge AI: Deploy AI models to edge devices with ONNX Runtime and IoT Hub.
  • 20 — Real-Time Analytics: Stream processing with AI-powered analysis using Event Hub and Stream Analytics.

Complexity Levels​

LevelPlaysDescription
Beginner08Minimal code, visual tools, guided setup
Intermediate01, 02, 03, 05, 06, 09, 10, 14, 15, 16, 17, 18Standard Azure + AI patterns, moderate config
Advanced04, 07, 11, 12, 13, 19, 20, 21, 22, 23Complex architecture, multi-service, deep expertise

Get Started with Any Play​

# 1. Browse plays
npx frootai list plays

# 2. Scaffold a play
npx frootai scaffold 01-enterprise-rag

# 3. Review the architecture
cat .github/copilot-instructions.md

# 4. Customize configuration
code config/openai.json config/guardrails.json

# 5. Deploy infrastructure (Azure plays)
az deployment group create --template-file infra/main.bicep \
  --parameters infra/parameters/dev.bicepparam

# 6. Use the builder agent
# In Copilot Chat: @builder Help me implement the core pipeline

:::tip Start with Play 01 If you're new to FrootAI, start with Play 01 — Enterprise RAG. It covers the most common AI pattern and teaches you the play structure. Then branch out to agents (Play 03/07), infrastructure (Play 02), or your specific use case. :::

Interactive Browser​

For the full interactive play browser with filtering, search, and detailed architecture diagrams, visit:

👉 frootai.dev/solution-plays