Enterprise AI Agent Deployment: Microsoft APM and Archestra
This article introduces Microsoft’s open‑source Agent Package Manager (APM) for installing, version‑locking, and publishing AI Agents, and the Archestra platform that provides multi‑model gateways, agent orchestration, MCP registration, security controls, RAG knowledge bases, and cost management for enterprise AI deployments.
Today we explore two popular open‑source projects in the Microsoft ecosystem that address the two major challenges of deploying AI Agents in enterprises: management and orchestration.
Microsoft APM – AI Agent Package Manager
Similar to npm, pip, or Homebrew, Microsoft APM (Agent Package Manager) is a package manager designed specifically for AI Agents.
It solves the emerging problem of installing, version‑managing, and avoiding dependency conflicts for the growing number of agents.
Install Agent : a single command pulls an agent from the registry.
Manage Versions : supports version locking and dependency declarations.
Publish Agent : developers can package and publish their own agents.
# Install an agent
apm install @microsoft/code-reviewer
# List installed agents
apm list
# Publish an agent
apm publish ./my-agentGitHub: https://github.com/microsoft/apm<br/>Stars: 2.7k+ | Language: Python | License: MIT
Archestra – Enterprise AI Platform
If APM handles "installation," Archestra handles "execution." Built in TypeScript, it supports Agent‑to‑Agent (A2A) communication and the MCP protocol, offering a comprehensive set of features:
Multi‑Model Gateway : a single entry point connects to ChatGPT, Claude, Gemini, DeepSeek, and other models.
Agent Orchestrator : enables agents to call each other, allowing complex tasks to be broken into collaborative agent workflows.
MCP Registry : manages tool registrations so agents can discover and invoke them dynamically.
Security Guardrails : enterprise‑grade permission controls, audit logs, and cost management.
Knowledge Base (RAG) : built‑in vector store supporting document upload and semantic retrieval.
Cost & Quota Management : per‑model and per‑user limits and budgeting, addressing key IT concerns.
# Quick start
npx archestra init
npx archestra startGitHub: https://github.com/archestra-ai/archestra<br/>Stars: 3.7k+ | Language: TypeScript | License: Apache-2.0
Both projects complement each other: APM manages distribution, while Archestra manages runtime, providing end‑to‑end coverage from installation to production deployment for AI Agents in enterprise settings.
Signed-in readers can open the original source through BestHub's protected redirect.
This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactand we will review it promptly.
Geek Labs
Daily shares of interesting GitHub open-source projects. AI tools, automation gems, technical tutorials, open-source inspiration.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.