What is {a}OS Explorer?

The interactive taxonomy and comparison engine for the agentic era.

{a}OS Explorer is a web application that lets you browse, compare, and understand AI products, frameworks, agents, workflows, and skills through a structured 7-layer reference model.

Think of it as the OSI model for the agentic era — but interactive.

New to the vocabulary? Use this translation first: Stratum = responsibility layer. Substrate = systems in that layer. Construct = thing they create or manage. Primitive = the smallest field, signal, or action inside it. The User Guide starts with a worked example so the vocabulary earns its keep instead of asking you to memorize jargon.

Key insight Most AI tooling sites tell you what a product claims to do. {a}OS Explorer tells you where it actually belongs, what it overlaps with, and what gaps appear when you compose systems together.

Why It Exists

The agentic AI ecosystem is growing fast. Hundreds of products, frameworks, and tools compete across overlapping domains. Architects, engineers, and technical leaders need a way to:

Understand where a product sits in the stack
See how products overlap or leave gaps
Compare Azure-native vs aftermarket solutions
Identify hidden dependencies between systems
Defend adoption decisions with evidence

Who It's For

User	Need
Architects	Compare stacks, identify overlap, assess governance implications
Engineers	Understand where tools sit and which layer is failing
CTOs / VPs	Explain architecture and defend adoption decisions
Procurement	Compare Azure-native vs aftermarket solutions
Security	Map compliance and governance controls across the stack

Key Features

🔍

Progressive Drilldown

Navigate Stratum → Substrate → Construct → Primitive. The UI unfolds — no page reloads.

⚖

Compare Basket

Add up to 4 products and see side-by-side coverage, overlaps, and gaps.

📊

Evidence-Backed

Every classification includes confidence scores and human-readable rationale.

🔀

Multi-Residency

Products can span multiple layers. The UI never forces false exclusivity.

🧬

Ontology-First

Structured data model beneath every view — not ad-hoc categories.

⌨

Keyboard-First

Full keyboard navigation. Press / to search, 1–7 to jump, ? for shortcuts.

Interactive Tools

🧠

Daily Stratum Quiz

One question per day — test which stratum a tool belongs to, see the rationale, and build a streak.

Open Quiz →

🧬

Stack Fingerprint

Generate a shareable image of your personal AI tool stack mapped across all 7 strata.

Generate Card →

Tech Stack

Layer	Technology
Framework	Next.js 15 + TypeScript
Styling	Tailwind CSS v4
Animation	Framer Motion + GSAP ScrollTrigger

The {a}OS Reference Stack

7 strata. 2 axes. From user intent down to raw model infrastructure.

The 7 Strata

Each stratum answers a specific boundary question about the agentic system:

Experience & Intent

What did the human actually want — and did they get it?

Governance & Trust

Who approved this action and under which policy?

Observability & Evaluation

How well did this work, and can we prove it?

Orchestration & Decisioning

What happens next, in what order, with which agent?

Execution & Interfaces

How does the agent act on the world?

Knowledge & Memory

What does the system know, and how does it recall?

Models & Infrastructure

What is the raw intelligence, and where does it run?

Cross-cutting Axes

Two axes cut across all strata, representing concerns that apply at every layer:

Axis	Scope
Governance & Trust	Policy enforcement, identity/access, compliance, audit, safety filters, content moderation
Observability & Evaluation	Distributed tracing, structured logging, benchmarks, A/B testing, drift/anomaly detection, SLO alerting

Ontology Hierarchy

Every AI product or tool is classified through a 4-level hierarchy:

Level	Definition	Example
Stratum	Top-level layer in the stack	L4 Orchestration & Decisioning
Substrate	Tools/frameworks operating in that stratum	Workflow Engines, Agent Frameworks
Construct	Meaningful artifact produced by a substrate	execution_plan, agent_roster
Primitive	Smallest atomic unit inside a construct	task_id, retry_count, priority

Agentic Concepts Glossary

Concept	Type	Definition
Hook	Mechanism	Event-driven callback that fires automatically at a lifecycle boundary (Stop, PostToolUse, SessionStart…). The runtime executes it — the model does not choose to call it.
Skill	Capability unit	Named, self-contained capability invoked on demand by a human (`/name`) or by an orchestrating agent. Always has a slash-command interface. The atomic unit of the {a}OS skill system.
Workflow	Orchestration	Multi-step sequence of skills, decisions, and human-confirmation gates that executes a complex goal. Workflows compose skills; skills do not compose workflows.
Heartbeat	Pattern	A recurring ambient check that runs quietly, surfaces findings to a clip file, and exits silently when clean. A pattern — not a mechanism. Implementable via hook, cron, or VS Code task.
Cron job	Mechanism	Time-based scheduler that fires at fixed intervals, independent of agent sessions. Use for daily briefs, weekly health scans, anything time-driven rather than event-driven.
VS Code Task	Mechanism	IDE-event-driven command defined in `.vscode/tasks.json`. Fires on folder open, build trigger, or keyboard shortcut. Use for pre-build lint, launch scripts, rename queues.

Rule of thumb: Hook / Cron / VS Code Task = how something fires (mechanisms). Heartbeat = what something does (pattern). Skill = what capability is available. Workflow = how capabilities chain.

Classification Glossary

Term	Meaning
Primary Stratum	The single layer where a product delivers its core value
Secondary Stratum	Additional layers a product touches with meaningful capability
Axis Role	A product's contribution to a cross-cutting concern (governance or observability)
Confidence Score	0.0–1.0 rating of how certain a classification placement is
Rationale	Human-readable explanation of why a product is placed at a given stratum
Multi-Residency	A product spanning multiple strata simultaneously

Classification Methodology

How products are mapped to the {a}OS reference model.

The Classification Model

Each entity in the explorer receives a structured classification:

Field	Description
Primary Stratum	The single layer where the product delivers its core value
Secondary Strata	Additional layers the product touches with meaningful capability
Axis Roles	Contributions to Governance & Trust or Observability & Evaluation
Capabilities	Named features mapped to specific strata
Constructs	Artifacts or state objects the product creates or manages
Primitives	Atomic units the product reads or writes
Confidence	0.0–1.0 certainty rating
Rationale	Explanation of the classification
Evidence	Optional links, quotes, or documentation

Confidence Levels

Level	Score	Badge	Meaning
High	≥ 0.85	0.92	Strong evidence, clear primary layer, multiple corroborating sources
Medium	0.60 – 0.84	~0.75	Reasonable placement, may have secondary interpretations
Low	< 0.60	~0.45	Disputed or emerging product, placement may change

Placement Rules

One primary stratum — every product gets exactly one core layer
Zero or more secondary strata — meaningful but non-primary capabilities
No forced exclusivity — a product spanning 5 layers shows all 5
Primary is visually stronger — full opacity vs dimmed for secondary
Evidence-backed — every placement has a rationale

Disputed Classifications

When a classification is uncertain:

Confidence score will be Medium or Low
Rationale will note the alternative interpretation
Future versions will support explicit dispute annotations

Example dispute A product classified as L4 (Orchestration) primary might note: "Could be argued as L3 (Execution) if tool-calling is considered the core loop rather than multi-step planning."

Product Classifications

The MVP product set classified against the {a}OS reference model.

MVP Product Set

Product	Vendor	Primary	Secondary	Deployment	Confidence
Azure AI Foundry	Microsoft	L1	L2, L3, L4	Cloud	0.92
Kotana	Internal	L4	L5, L6, L7	Hybrid	0.88
Paperclip	Internal	L7	L3, L4	Local	~0.75
CrewAI	CrewAI Inc	L4	L3	Cloud	0.90
Mem0	Mem0 AI	L2	—	Cloud	0.95
LangGraph	LangChain	L4	L3	Hybrid	0.88
Azure AI Search	Microsoft	L2	—	Cloud	0.93
Open Policy Agent	CNCF	L6	—	Hybrid	0.95

Azure AI Foundry

Primary: L1 Models & Infrastructure 0.92

Secondary: L2 Knowledge, L3 Execution, L4 Orchestration

Rationale: Built on Microsoft.CognitiveServices. Model hosting and inference is the core value prop; prompt flow, search, and agent capabilities are secondary strata.

Kotana

Primary: L4 Orchestration & Decisioning 0.88

Secondary: L5 Observability, L6 Governance, L7 Experience

Rationale: Kotana is an orchestration-first platform. Governance, observability, and UX layers are secondary capabilities.

Paperclip

Primary: L7 Experience & Intent ~0.75

Secondary: L3 Execution, L4 Orchestration

Rationale: Desktop copilot with strong UX and intent parsing. Orchestration and execution are secondary.

CrewAI

Primary: L4 Orchestration & Decisioning 0.90

Secondary: L3 Execution

Rationale: Multi-agent crew framework focused on orchestrating agent collaboration and task delegation.

Mem0

Primary: L2 Knowledge & Memory 0.95

Secondary: None

Rationale: Purpose-built memory layer for AI. Manages context persistence, recall, and long-term agent memory.

LangGraph

Primary: L4 Orchestration & Decisioning 0.88

Secondary: L3 Execution

Rationale: Stateful graph-based agent orchestration framework from LangChain. Defines execution flows as directed graphs with state management.

Azure AI Search

Primary: L2 Knowledge & Memory 0.93

Secondary: None

Rationale: Cloud-native search-as-a-service with vector + hybrid retrieval. Core of many RAG pipelines.

Open Policy Agent

Primary: L6 Governance & Trust 0.95

Secondary: None

Rationale: General-purpose policy engine (Rego-based). Evaluates access control, resource constraints, and compliance policies.

User Guide

How to get the most out of {a}OS Explorer.

Terminology Primer

Five nouns power the entire {a}OS model. Lock these in and the Explorer feels intuitive instead of academic.

Plain-English shorthand — Every official term has a plain alias in parentheses. Use whichever clicks. Both appear throughout the Explorer UI.

You see…	We call it	What it means	Explorer example
Pick a layer	Stratum (layer)	One of 7 horizontal tiers in the stack, each defined by a boundary question — “Who decides?”, “How does it learn?”, etc.	L4 Orchestration & Decisioning
Notice a theme that cuts across layers	Axis (cross-cut)	A concern that spans multiple strata vertically — security, cost, latency, compliance, etc. Axes surface in every layer but are governed holistically.	Security axis — touches L6 Governance, L3 Execution, and L1 Infrastructure
Expand a group inside a layer	Substrate (group)	A named group of related capabilities within a stratum. Each substrate bundles the frameworks, engines, or services that do work at that level.	Workflow Engines or Agent Frameworks inside L4
Click a card	Construct (concept / card)	A named concept or data structure within a substrate. Constructs are the items you browse, compare, and classify in the Explorer.	`execution_plan`, `agent_roster`
Read the fields on a card	Primitive (field)	The smallest atomic part inside a construct — a field, flag, token, parameter, or actionable unit.	`task_id`, `retry_count`, `priority`

Optional mental image — If it helps, picture the stack as a vertical city: each stratum is a floor, substrates are the departments on that floor, constructs are the deliverables they produce, and primitives are the parts inside each deliverable.

Terminology Infographic

🖼 Place {a}OS Terminology Primer infographic here
Save the image to 90.aOS-Explorer/docs/ and update the <img> tag

Quick Start

Open the Explorer — you'll see the 7-stratum reference stack
The guided tour will introduce the key areas
Click any stratum to expand its substrates
Click a substrate to see its constructs
Click product cards (right panel) to see classification details

Browsing the Stack

The central canvas shows the {a}OS stack — 7 layers stacked vertically from L7 (top, user-facing) to L1 (bottom, infrastructure).

Each layer card shows:

A color-coded bar on the left edge
A layer number (L1–L7)
A name and boundary question
A substrate count

Click any layer to expand it in-place. Inside you'll find substrates, each with their own constructs. The breadcrumb at the top tracks your depth: {a}OS Stack → L4 Orchestration → Workflow Engines

Comparing Products

Click the + button on any product card to add it to your compare basket (up to 4). The badge in the header shows your count.

The compare view provides:

Layer coverage heatmap
Axis role comparison
Construct/primitive overlap table
Gap and overlap summary

Filtering

Discoverability tip The top mode pills and hero chips mirror the left rail. If you click a quick chip like Frameworks, MCPs, or Open Source, the matching sidebar filter updates so you can learn what the full filter system supports.

The left rail provides filter groups:

Filter	Values
Type	Products, Frameworks, Workflows, Agents, Skills, MCPs
Vendor	Microsoft, CrewAI Inc, Mem0 AI, CNCF, etc.
Deployment	Cloud, Local, Hybrid
License	Open Source, Proprietary
Stratum	L1–L7
Substrate	Agent Frameworks, Workflow Engines, Policy Engines, etc.
Construct	execution_plan, agent_roster, policy_bundle, memory_index

Keyboard Shortcuts

Key	Action
`/`	Focus search bar
`1` – `7`	Jump to stratum L1–L7
`↑` `↓`	Navigate between strata
`→`	Expand / drill deeper
`←`	Collapse / drill up
`c`	Toggle compare mode
`?`	Show all shortcuts
`Esc`	Close / collapse all

Interaction Modes

Mode	Description
Stack Mode	Classic 7-stratum vertical view (default)
Accordion Mode	Expand strata into substrate/construct/primitive rows
Compare Mode	Side-by-side comparison of 2–4 selected items
Category Mode	Toggle between Products, Frameworks, Agents, Workflows, Skills
Map Mode	Multi-layer overlap visualization
Axis Mode	Cross-cutting governance/observability analysis

Roadmap

Where {a}OS Explorer is today and where it's headed.

Current — MVP

✅ 7-stratum interactive explorer with progressive drilldown
✅ Stratum → Substrate → Construct expansion
✅ 5 featured product classifications with rationale
✅ Compare basket (add up to 4 products)
✅ Left rail: strata nav, axis toggles, filter placeholders
✅ Right rail: product cards, detail panel
✅ Full keyboard navigation
✅ Guided onboarding tour
✅ Cinematic dark-theme UI with aurora effects

Next Phase

🚧 Full compare screen with heatmap and gap analysis
🚧 Dedicated product detail pages
🚧 Dedicated stratum detail pages
🚧 Connected fuzzy search across all entities
🚧 Active filter system (vendor, deployment, license, maturity)
🚧 Framework comparison with translation matrix
🚧 Evidence panel with source links and quotes
🚧 All 8 MVP products fully classified
🚧 Mode pills connected to canvas views
🚧 Primitive-level drill (Construct → Primitives)

Future

💡 Saved comparison sets
💡 AI-assisted classification suggestions
💡 Team workspaces and collaboration
💡 Public submission pipeline with moderation
💡 Scenario-based evaluation mode
💡 Framework scoring and weighting system
💡 Alternate framework overlays

FAQ

Is this a product directory?

No. It's a classification and comparison engine. Products are mapped to a structured ontology, not just listed.

Can I add my own products?

Not in the MVP. Future phases will support submission pipelines with moderation.

How is {a}OS different from other AI frameworks?

{a}OS provides 7 distinct strata with explicit boundary questions, multi-residency classification, and cross-cutting axes. Most alternatives collapse these layers into fewer, coarser categories.

What if I disagree with a classification?

Every placement shows its confidence score and rationale. Disputed classifications are explicitly marked. Future versions will support formal dispute annotations.

What tech stack does this use?

Next.js 15, TypeScript, Tailwind CSS v4, with Aceternity UI, Framer Motion, GSAP, and Zustand. Data starts as structured JSON/YAML and evolves to Prisma + Postgres.

Azure AI Foundry Setup

Step-by-step guides for standing up Azure AI Foundry, organized by {a}OS stratum. Each section shows what Foundry provides, what you must build, and how to set it up.

Glossary — Abbreviations Used Below

GA: Generally Available (production-ready)
Preview: Public preview (not SLA-backed)
PTU: Provisioned Throughput Units (reserved capacity)
CMK: Customer-Managed Keys (encryption)
SFT: Supervised Fine-Tuning
DPO: Direct Preference Optimization
RFT: Reinforcement Fine-Tuning
RAG: Retrieval-Augmented Generation
BYO: Bring Your Own (storage / resources)
MCP: Model Context Protocol
RBAC: Role-Based Access Control
APIM: Azure API Management
VNet: Virtual Network (Azure networking)
IaC: Infrastructure as Code
PII: Personally Identifiable Information
PaaS: Platform as a Service
FIPS: Federal Information Processing Standards
Entra ID: Microsoft's identity platform (formerly Azure AD)
Bicep: Azure's declarative IaC language
Rego: Policy language for Open Policy Agent

Getting Started

Azure AI Foundry (rebranded to "Microsoft Foundry" in 2026) is Microsoft's unified PaaS for AI — built on Microsoft.CognitiveServices. It bundles model hosting, agent frameworks, content safety, evaluation, and a management portal into a single control plane.

Two resource models coexist The "classic" model (based on Microsoft.MachineLearningServices) and the "new" model (based on Microsoft.CognitiveServices) are both active. The new model is where Microsoft invests, but the classic model has more mature network isolation. Check which you are on: az provider show -n Microsoft.CognitiveServices

Prerequisites:

Azure subscription (pay-as-you-go is fine for learning)
Azure CLI installed — az --version
Python 3.10+ and pip install azure-ai-projects azure-identity
Register the provider: az provider register -n Microsoft.CognitiveServices

Foundry coverage across the {a}OS stack:

Stratum	Coverage	Key Takeaway
L7 Experience	Partial	Portal + publishing, but new portal lacks VNet support
L6 Governance	Partial	Content Safety + RBAC. No approval gates, no cost governance
L5 Observability	Partial	Tracing + evaluators. Breaks with private App Insights
L4 Orchestration	Partial	Prompt agents GA, workflow agents preview
L3 Execution	Good	Function calling, MCP, Code Interpreter all GA
L2 Knowledge	Partial	AI Search GA, Foundry IQ preview. No cross-agent state
L1 Models	Strong	Model catalog, PTU, fine-tuning, CMK — all GA

7-Day Learning Path

A structured curriculum that builds bottom-up through the strata. Each day: ~2-4 hours. See the full curriculum for detailed instructions.

L1 — Models

Create Foundry resource, deploy model, first inference. Posts 1-2.

L2+L3 — Data & Tools

RAG pipeline, AI Search, tool calling, MCP. Posts 5, 8.

L4 — Agents

Prompt agent, identity, versioning, maturity matrix. Post 7.

L5 — Observability

App Insights, evaluators, CI/CD gates, private tracing gap. Post 10.

L6+L7 — Governance

RBAC hardening, Content Safety, approval architecture. Posts 3, 4, 6.

Cross-Cutting

Network isolation, VNet injection, APIM cost governance. Posts 5, 9.

Integration

Bicep deployment, gap analysis, {a}OS classification. Posts 11, 12.

L1: Models & Infrastructure

Models & Infrastructure

Is the model/runtime/compute/network foundation available and behaving correctly?

What Foundry Provides

Capability	Feature	Status
Model hosting	Model catalog (11K+ models, ~50-100 that matter)	GA
Deployment types	Standard, PTU, Global, DataZone, Batch	GA
Fine-tuning	SFT/DPO for GPT-4o/4.1; RFT for o4-mini	GA
Encryption	CMK via Azure Key Vault (FIPS 140-2)	GA
Edge inference	Foundry Local	GA

What You Build

Custom model training (requires Azure ML — separate service)
Cross-region disaster recovery (manual architecture)
Gov cloud model gap mitigation (model abstraction layer)
Hybrid cloud patterns for classified environments

Setup Walkthrough

Create a Foundry resource: az cognitiveservices account create --kind AIServices --sku S0
Create a project within the resource
Deploy a model: az cognitiveservices account deployment create --model-name gpt-4.1 --sku-name GlobalStandard
Test inference via Python SDK or REST API
Enable CMK encryption via Key Vault if handling sensitive data

Key Decisions

Decision	Option A	Option B	Recommendation
Deployment type	Pay-as-you-go	PTU (reserved)	Start with pay-as-you-go; PTU when you have predictable load
Scope	Global deployment	Regional	Regional for data residency; Global for cost optimization
Model portability	Hardcode model names	Abstraction layer	Always abstract — build for the model you might lose access to

Related: LinkedIn Post 2 (What Foundry Is), Post 6 (Models & Money) — Learning Day 1

L2: Knowledge & Memory

Knowledge & Memory

What context did the system know, retrieve, remember, or forget?

What Foundry Provides

Capability	Feature	Status
RAG engine	Foundry IQ	Preview
Vector search	Azure AI Search	GA
Agent memory	Conversation memory per agent	GA
File processing	File Search tool	GA (no VNet)
Document processing	Document Intelligence	GA

What You Build

Cross-agent shared state (Cosmos DB)
Memory compaction and lifecycle management
Production RAG governance and access controls
Context window management strategies

Critical: Basic vs. Standard Storage "Basic" agent storage is Microsoft-managed multitenant. If you handle sensitive data, PII, financial records, or any controlled data, you need "Standard" with BYO storage — and you need to know this before your first deployment, not after.

Setup Walkthrough

Create Azure AI Search: az search service create --sku standard
Configure private endpoint for AI Search
Upload documents and create a vector index
Connect AI Search as an agent tool
Configure BYO storage (Standard agent setup) — do not use Basic for production

Related: LinkedIn Post 8 (RAG & Data) — Learning Day 2

L3: Execution & Interfaces

Execution & Interfaces

What external action was invoked, against which interface, and what happened?

What Foundry Provides

Capability	Feature	Status	VNet
Function calling	Native function/tool calling	GA	Works
MCP support	MCP server integration	GA	Works
Code execution	Code Interpreter	GA	Works
Enterprise connectors	1,400+ via Logic Apps	GA	Partial
Web search	Bing Grounding	GA	PUBLIC
Browser	Browser Automation	Preview	Broken

Warning: Public Internet Tools Bing Grounding and Web Search route over public internet even in "isolated" environments. Block via Azure Policy on day one:

az policy assignment create --name block-bing --policy "deny Bing
              grounding"

Setup Walkthrough

Add a function tool to your agent — test with a simple calculator function
Set up MCP server integration for your own services
Test Code Interpreter with a data analysis task
Block Bing/Web Search via Azure Policy in production
Build your personal VNet tool compatibility checklist

Related: LinkedIn Post 5 (Network Isolation), Post 7 (Agent Maturity) — Learning Day 2

L4: Orchestration & Decisioning

Orchestration & Decisioning

What happens next, in what order, with which agent, under which stop conditions?

What Foundry Provides

Agent Type	Status	VNet	Production Ready
Prompt Agents	GA	Yes	Yes
Workflow Agents	Preview	Partial	No
Hosted Agents	Preview	No	No

Start with Prompt Agents They are GA and production-safe. Workflow and hosted agents are preview — build your first value on what is stable, not what is exciting. Complex orchestration? Use Semantic Kernel or LangGraph in code.

Setup Walkthrough

Create a prompt agent with a detailed system prompt
Attach tools from L3 (function calling, MCP)
Configure per-agent Entra managed identity
Test agent in the playground
Create a second version — test versioning and rollback
Explore workflow agents (preview) — understand their limitations before committing

Related: LinkedIn Post 7 (Agent Maturity) — Learning Day 3

L5: Observability & Evaluation

Observability & Evaluation

What actually happened, how well did it perform, and where did it fail?

What Foundry Provides

Capability	Feature	Status
Tracing	OpenTelemetry via App Insights	GA (prompt agents)
Evaluators	Quality, safety, task adherence	GA
Custom evaluation	Configurable pipelines	GA
Continuous eval	Scheduled eval against live data	GA
CI/CD	GitHub Actions evaluation gates	GA

Critical Gap: Private App Insights Tracing does NOT work with private Application Insights. If you need both network isolation and observability — which is most serious deployments — you must architect a workaround: non-private App Insights in a peered network with restricted access, or export traces to Log Analytics.

Setup Walkthrough

Create Application Insights: az monitor app-insights component create
Connect to your Foundry project
Run your agent — inspect traces in the portal
Run a quality evaluator against agent output
Run a safety evaluator — verify Content Safety scoring
Set up a GitHub Actions workflow with an evaluation gate
If using private networking: test tracing (it may break) and plan the workaround

Related: LinkedIn Post 10 (Observability Paradox) — Learning Day 4

L6: Governance & Trust

Governance & Trust

Who is allowed to do this, under what policy, with what risk and budget?

What Foundry Provides

Capability	Feature	Status
Content filtering	Content Safety (always on, configurable)	GA
Prompt protection	Prompt Shields — jailbreak detection	GA
RBAC	Entra ID + 4 built-in roles + custom	GA
Infra policy	Azure Policy integration	GA
Threat detection	Microsoft Defender	GA

What You Build (This Is the Big One)

Governance is not a feature — it's an architecture Foundry has no approval workflows, no governance dashboard, no aggregated audit trail, no compliance reporting, no agent behavior policy engine, and no cost governance. Everything in this list is your responsibility.

RBAC Roles

Role	Use For	Risk Level
Azure AI User	Developers (least privilege)	Low — default for all devs
Azure AI Project Manager	Team leads	Medium
Azure AI Account Owner	Resource creation only	Medium — no data actions
Azure AI Owner	Never without justification	High — loaded gun in prod

Setup Walkthrough

Switch to Entra ID authentication — remove all API keys
Assign Azure AI User to developers (least privilege)
Verify Azure AI Owner is NOT assigned to any developer accounts
Configure Content Safety severity thresholds
Test Prompt Shields — attempt a jailbreak, verify detection
Create Azure Policy assignments to block dangerous tools in production
Design your approval gate architecture (Logic Apps or custom workflow — Foundry does not provide this)

Related: LinkedIn Post 3 ({a}OS debut), Post 4 (Identity) — Learning Day 5

L7: Experience & Intent

Experience & Intent

What is the user asking, seeing, approving, or rejecting?

What Foundry Provides

Capability	Feature	Status	Limitation
Portal	Foundry Portal (classic + new)	GA	New portal lacks VNet support
Agent builder	Visual agent configuration	GA	—
Publishing	Teams, M365, BizChat	GA	Requires public endpoints
API surface	REST + Python/JS SDKs	GA	—
Playground	Agent testing playground	GA	—

Most of L7 is yours to build The Foundry Portal is a management surface, not an end-user experience. For most production deployments, the L7 surface — the thing your actual users see and interact with — will be entirely custom-built. Agent publishing to Teams requires public endpoints, which is often a non-starter for security-conscious environments.

Setup Walkthrough

Navigate the Foundry Portal — toggle between classic and new portals
Use the classic portal or CLI for network-isolated management (the new portal doesn't support it)
Test agent publishing to Teams (note the public endpoint requirement)
Design your L7 surface: what will end users actually see?
Plan: approval surfaces, escalation flows, intent disambiguation — none of these come from Foundry

Related: LinkedIn Post 3 (Governance debut) — Learning Day 5

Network Isolation

Three isolation dimensions: inbound (private endpoints), outbound from resource (VNet injection), and outbound from agents (tool-level).

What Works in VNet

Tool / Feature	VNet Status
Private endpoints (inbound)	Works
VNet injection (agent compute)	Works (/27+ subnet)
Function calling	Works
MCP (private)	Works
Code Interpreter	Works
AI Search (private endpoint)	Works

What Breaks in VNet

Tool / Feature	Issue
File Search	Not supported in VNet isolation
OpenAPI Tool	Not supported in VNet isolation
Azure Functions (tool)	Not supported in VNet isolation
Browser Automation	Not supported in VNet isolation
Hosted Agents	No VNet support at all
Workflow Agents	Inbound only
Tracing (private App Insights)	Breaks entirely
New Foundry Portal	Must use classic, SDK, or CLI

What Goes Over Public Internet

Tool	Risk
Bing Grounding	Public internet even in "isolated" environments
Web Search	Public internet even in "isolated" environments
SharePoint Grounding	Public internet even in "isolated" environments

Block these via Azure Policy in production The marketing says "end-to-end network isolation." The reality is an exceptions list. Know it before your security team finds it for you.

Related: LinkedIn Post 5 (Network Isolation Exceptions Table) — Learning Day 6

Cost Governance

The platform itself is free to explore. Costs come from model consumption, storage, search, networking, and compute.

Native vs. Custom

Capability	Source
Budget alerts	Native — Azure Cost Management
Per-resource tracking	Native — Azure Cost Management
Per-team chargeback	Custom — requires APIM gateway
Per-user token budgets	Custom — requires APIM + logic
Real-time spend enforcement	Custom — requires APIM kill switch
Cross-service attribution	Custom — Log Analytics queries

Deploy APIM as an AI Gateway from Day One Azure API Management sits between your consumers and Foundry. It provides token metering, rate limiting, and per-team chargeback that Foundry lacks natively. Build it before you need it — not after an agent loops and spends your monthly budget overnight.

Related: LinkedIn Post 9 (Cost Governance) — Learning Day 6

IaC & CI/CD

Deploy via Bicep/CLI, not the portal. The new portal does not support network isolation workflows. IaC is not optional — it's the only reliable path for production.

Bicep Templates

00-basic — Simple Foundry resource + project
19-hybrid-private-resources-agent-setup — Network-isolated agent setup with BYO storage

Fork the azure-ai-foundry/foundry-samples Bicep templates. Customize for your policies and compliance. Maintain as an internal module.

Phase 1 Architecture (Minimum Viable Platform)

Foundry resource + project
Private endpoints
One model deployment (pay-as-you-go)
One prompt agent (GA)
Application Insights (with tracing workaround if private)
APIM AI Gateway (cost metering from day one)
Azure Cost Management budget with alerts
Entra-only authentication (no API keys)

CI/CD Pattern

Bicep deployment via GitHub Actions
Agent evaluation gate (quality + safety evaluators)
Cost check gate (budget threshold)
Promote: dev → staging → prod (separate Foundry resources per environment)

Projects are NOT environments Projects within a single Foundry resource share networking and cannot serve as environment boundaries. Use separate Foundry resources per environment (dev, staging, prod).

Related: LinkedIn Post 11 (Phase 1) — Learning Day 7