Skip to main content

🎯 Solution Plays

Each play ships with the full .github Agentic OS (19 files, 4 layers), 🛠️ DevKit (empower your coding agent), 🎛️ TuneKit (fine-tune AI for production), and plugin.json (Layer 4 distribution manifest). LEGO blocks that compose into complete solutions.

🧠

.github Agentic OS

Layer 1 — Instructions (always-on context)
Layer 2 — Prompts, Agents, Skills
Layer 3 — Hooks + Workflows (CI/CD)
Layer 4 — Plugin packaging

19 files per play · 7 primitives · 4 layers

🛠️

DevKit — Build + Deploy

agent.md — solution-aware co-coder
instructions — prompts, guardrails
infra/ — Bicep IaC templates
MCP + plugins — tools + functions

Code + infrastructure in one kit.

🎛️

TuneKit — AI Fine-Tuning

config/*.json — temp, top-k, models
agents.json — agent behavior tuning
model-comparison — cost vs quality
evaluation/ — test + score + ship

Tune AI without being an AI specialist.

3 ready · 17 skeleton · 20 total

🔍 01 — Enterprise RAG Q&A

ReadyMedium

Production RAG — AI Search + OpenAI + Container Apps. Pre-tuned: temp=0.1, hybrid 60/40, top-k=5, semantic reranker.

Infra: AI Search · Azure OpenAI · Container Apps · Blob
Tuning: temperature · top-k · chunk size · reranking · relevance threshold

⛰️ 02 — AI Landing Zone

ReadyFoundation

Foundational Azure infra for AI — VNet, private endpoints, managed identity, RBAC, GPU quotas.

Infra: VNet · Private Endpoints · RBAC · Managed Identity · Key Vault
Tuning: Network config · service SKUs · GPU quota · region

🎯 03 — Deterministic Agent

ReadyMedium

Reliable agent — temp=0, structured JSON, multi-layer guardrails, anti-sycophancy, evaluation.

Infra: Container Apps · Azure OpenAI · Content Safety
Tuning: temperature=0 · JSON schema · seed · confidence · citations

📞 04 — Call Center Voice AI

SkeletonHigh

Voice-enabled customer service — Communication Services + AI Speech + OpenAI Agent.

Infra: Communication Services · AI Speech · Azure OpenAI · Container Apps
Tuning: Speech config · grounding prompts · fallback chains · voice personality

🎫 05 — IT Ticket Resolution

SkeletonMedium

Auto-classify, route, and resolve IT tickets — Logic Apps + AI Foundry + ServiceNow MCP.

Infra: Logic Apps · Azure OpenAI · ServiceNow MCP · Container Apps
Tuning: Classification prompts · routing rules · confidence thresholds

📄 06 — Document Intelligence

SkeletonMedium

Extract, classify, and structure document data — Blob + Doc Intelligence + OpenAI.

Infra: Blob Storage · Document Intelligence · Azure OpenAI · Cosmos DB
Tuning: Extraction prompts · confidence thresholds · field schemas

🤖 07 — Multi-Agent Service

SkeletonHigh

Supervisor agent routes to specialists — Container Apps + Agent Framework + Dapr.

Infra: Container Apps · Azure OpenAI · Cosmos DB · Dapr
Tuning: Supervisor routing · tool schemas · agent memory · handoff rules

💬 08 — Copilot Studio Bot

SkeletonLow

Low-code enterprise bot — Copilot Studio + Dataverse + SharePoint.

Infra: Copilot Studio · Dataverse · SharePoint · Power Platform
Tuning: Topic design · knowledge sources · guardrails · flow

🔎 09 — AI Search Portal

SkeletonMedium

Enterprise search — AI Search + Web App + Semantic Ranking.

Infra: AI Search · App Service · Azure OpenAI · Blob Storage
Tuning: Hybrid weights · semantic config · scoring profiles · filters

🛡️ 10 — Content Moderation

SkeletonLow

Filter harmful content — AI Content Safety + APIM + Functions.

Infra: Content Safety · API Management · Azure Functions
Tuning: Severity levels · custom categories · blocklists · actions

🏔️ 11 — Landing Zone Advanced

SkeletonHigh

Multi-region, policy-driven, enterprise-grade AI landing zone with governance at scale.

Infra: Multi-region VNet · Azure Policy · PE · RBAC · GPU Quota
Tuning: Governance policies · multi-region config · advanced RBAC

⚙️ 12 — Model Serving AKS

SkeletonHigh

Deploy and serve LLMs on AKS with vLLM, GPU nodes, and auto-scaling.

Infra: AKS · vLLM · NVIDIA GPU · Container Registry
Tuning: Quantization · batching · scaling rules · model config

🔬 13 — Fine-Tuning Workflow

SkeletonHigh

End-to-end fine-tuning — data prep, LoRA training, evaluation, deployment.

Infra: AI Foundry · GPU Compute · Storage · MLflow
Tuning: LoRA rank · learning rate · epochs · eval metrics

🚪 14 — AI Gateway

SkeletonMedium

APIM-based AI gateway with semantic caching, load balancing, and token budgets.

Infra: API Management · Redis Cache · Azure OpenAI (multi-region)
Tuning: Token budgets · caching rules · fallback chain · rate limits

🖼️ 15 — Multi-Modal DocProc

SkeletonMedium

Process documents with text + images using GPT-4o multi-modal capabilities.

Infra: GPT-4o · Blob Storage · Cosmos DB · Functions
Tuning: Image prompts · extraction schemas · confidence thresholds

👥 16 — Copilot Teams Extension

SkeletonMedium

M365 Copilot extension with Graph API and declarative agents.

Infra: M365 Copilot · Graph API · Azure Functions · App Reg
Tuning: Declarative agent config · permissions · scoping

📊 17 — AI Observability

SkeletonMedium

Monitor AI workloads with KQL queries, quality alerts, and workbooks.

Infra: App Insights · Log Analytics · Azure Monitor · Workbooks
Tuning: KQL queries · alert thresholds · quality metrics

📝 18 — Prompt Management

SkeletonMedium

Version control, A/B test, and rollback prompts across environments.

Infra: Prompt Flow · Git · GitHub Actions · AI Foundry
Tuning: Prompt versions · A/B weights · rollback rules

📱 19 — Edge AI Phi-4

SkeletonHigh

Deploy Phi-4 on edge devices with ONNX quantization and local serving.

Infra: IoT Hub · Container Instances · ONNX Runtime · Edge
Tuning: Quantization level · model config · sync schedule

🚨 20 — Anomaly Detection

SkeletonHigh

Detect anomalies in real-time streams — Event Hub + AI analysis.

Infra: Event Hub · Stream Analytics · Azure OpenAI · Functions
Tuning: Threshold config · alert prompts · detection windows
🔗 Back to Ecosystem🌳 Back to FrootAI