Editorial
As the internet has evolved over the past few decades, including overhauls to meet mobile consumption and scaling service to the cloud, agentic computing is the next great flashpoint.
What’s developing is an IT landscape of numerous AI agents, or autonomous software entities capable of reasoning, collaborating and executing autonomous tasks, fully reshaping back-end and front-end interactions.
On the back-end, AI agents embed into an IT platform, sitting above a digital ecosystem. They are trained to manage deployments, configuration drift, service level agreements (SLA) compliance and more. On the front-end, agents can execute consumer-facing tasks like directly managing customer service calls, booking flights, reordering diapers and more. These are not chatbots, but an army of AI agents trained to reason, collaborate with other agents and take action.
Table of Contents
- Why Agentic AI Breaks Without a Composable Core
- Where Monoliths Start to Crack
- How Composable Better Supports Agentic
- Scaling Agents Requires More Than Models
Why Agentic AI Breaks Without a Composable Core
Early agentic activity is happening today online. But to scale it and further adopt agentic innovation, enterprise businesses must construct a flexible foundation.
To do so, companies must migrate away from monolithic frameworks and push toward composable architectures.
The reason is simple: AI agents don’t work as an army of one, rather a force of multiple agents from several sources working in concert. Interoperable software empowers internet agents to work together, learn from one another and access unified and shared data to make personalized and business-specific decisions.
Related Article: How to Build Multi-Agent Workflows That Don't Fall Apart
Where Monoliths Start to Crack
Currently, many IT departments feel constrained by headcount, technical skills and bandwidth. And while agentic computing aims to directly solve for some of this, adding in the technology comes with its own complexities.
From a people perspective, there’s training and knowledge to be done, so that agents are supported and acting accurately on behalf of the business.
There are also technical requirements, and traditional, monolithic architectures and platforms can crack under the pressure. In a monolith, siloed data often prevents agents from building accurate contextual understandings, while tightly coupled logic can block agents from adapting workflows. Closed ecosystems, by their nature, create vendor lock-in and make cross-agent collaboration impossible.
For agentic systems to work effectively and be scalable, open pathways for agent-to-agent communication must be in place. MACH principles and composable architecture support agentic implementations through:
- API-first, modernized microservices that make capabilities consumable by autonomous agents.
- Standardized open APIs that ensure interoperability across internal and external agent ecosystems.
- Agent orchestration that is governed, observed and provides total data access and secure engagements.
- Flexible design in which agents work together and new ones emerge and are integrated seamlessly without replatforming.
How Composable Better Supports Agentic
Research from the MACH Alliance shows that organizations advanced in their composable architectures report 77% AI usage, compared to only 36% of those new to MACH and composable principles.
Inside easyJet’s Agentic Roadmap
Again, these AI experiences uplevel front-end and back-end experiences. A company like easyJet, a leading European airline, is finding success in the modularity and API-first design of composable infrastructures, enabling better agent-to-agent communication.
The airline, a member of the MACH Alliance, is using AI agents to manage bookings, customer service prompts and internal operations for faster workflows. The company anticipates by 2030 that 60% of all its online transactions will be initiated through AI agents. For the company, a MACH-based approach acts as a bridge between internet computing and AI. Flexible design, modular components and composable technology provide stability, while preparing companies for future innovation.
Designing a Retail-Grade MCP Architecture
Similarly, a sporting goods retailer, let’s call it Sportz, can use composable principles to develop an autonomous, 24/7 digital salesperson with perfect memory and access to the brand’s entire knowledge graph. Sportz would put in place a multi-agent system powered by a range of model context protocol (MCP) servers that sync everything from product data to inventory, pricing, PIM attributes, customer preferences, loyalty balances and fulfillment logic.
The ecosystem would plug agentic into:
- Product MCP servers (PIM, catalog, multi-storefront configuration)
- Customer MCP servers (profile, segmentation, preferences, historical behavior)
- Commerce MCP servers (cart, checkout, payments, promotions, loyalty, returns)
- Content MCP servers (CMS, blog, PDP/PLP content blocks)
- Ops MCP servers (OMS, delivery ETA, inventory per supply channel)
- Internal knowledge MCP servers (policies, sizing guides, warranty rules, staff training docs)
- External MCP servers (weather, events, geolocation, travel schedules, brand data, influencer feeds)
The result: an online operation orchestrated by a highly tuned model that understands Sportz’s tone, rules and merchandising strategies.
Related Article: Inside Anthropic’s Model Context Protocol (MCP): The New AI Data Standard
Scaling Agents Requires More Than Models
AI agents are transforming the internet, commerce and operations, but sustaining this monumental change means implementing digital frameworks that are modular and interoperable.
Webinar
Jan
29
Cut the Noise: Deploying AI That Actually Moves the Needle
Learn how to turn AI experimentation into concrete revenue operations.
Register
Webinar
On demand
Rebrand. Migrate. Optimize. How to Do It All (Without Slowing Down)
Cresta leveled up site speed, design flexibility and marketer sanity (in record time). Find out how.
Watch Now
Webinar
On demand
From Manual to Magical: How AI Transforms CX Teams
Learn how to replace manual support processes with automation that actually delivers.
Watch Now
Webinar
On demand
How to Build a Solid Knowledge Foundation for AI Success
See how leading brands keep their AI honest, compliant and actually helpful.
Watch Now
Webinar
On demand
Fix the Content Bottleneck: Build a Better WebOps Strategy
Content stalled? Dev overloaded? You’re not the only one. Learn how streamlined WebOps bridges the publishing gap.
Watch Now
Webinar
On demand
Beyond Storage: Smarter Content, Bigger Impact with DAM + AI
Discover how the DAM + AI duo makes content smarter, stronger and more accessible.
Watch Now
Monolithic frameworks cannot help companies grow along with agentic computing. AI agents require a flexible foundation, enabling multiple agents to collaborate, leveraging shared, governed data and coordinated inside a composable architecture.
Companies won’t reach the potential of agentic using monolithic platforms that operate on isolated performers. Companies need to develop platforms guided by MACH, and they need to continue to build and share education that supports clearer understandings of how composability and agentic work together.
The rise of agentic computing is a major moment, but companies need to embrace it in the right way to reach its full potential.
Learn how you can join our contributor community.
