AI Agents Are Now Fixing the Broken Middle of Your Business — The June 11 Stories That Prove It

“The most expensive part of most businesses is not their people or their tools. It is the gap between them. The copy-paste workflows. The emails that bridge systems that never talked. The manual handoffs between departments should be automatic. That is what AI agents are now being built to fix — and June 11, 2026 delivered two stories that show exactly how.”

The Two Stories Breaking Today

Two announcements published this morning tell the same underlying story from different angles. One comes from Publicis Sapient, a global enterprise AI services company. The other comes from a developer working with Claude Opus 4.8’s dynamic workflow capability. Together, they describe where AI automation is actually landing in 2026: not in flashy demos, but in the broken middle of how organisations operate.

Story One: Publicis Sapient Launches Sustain — Agentic AI for IT Operations

What Was Announced

Publicis Sapient today launched Sapient Sustain, an agentic AI platform designed to transform IT operations and managed services. The product targets what Publicis calls “the turf war between enterprise AI modernisation and legacy systems separated by time, data, distribution, and integration.”

The problem Sustain addresses is one I see in almost every business I work with at Hexona Systems, regardless of size: employees spending significant time copying and pasting information between systems, entering data into spreadsheets that duplicate information already in another platform, emailing back and forth to bridge gaps between departments, and operating with no visibility across the organisation because no system talks to any other.

The result, as Publicis describes it, is a “tangled web of communication between different departments, a lack of visibility within the same organisation, and inconsistent employee and customer experiences.” That description fits the operational reality of most businesses running more than five software tools.

What Sustain Actually Does

Sustain deploys AI agents specifically into the IT operations and managed services layer. The agents handle:

• Incident detection and automated first-response triage
• Cross-system visibility that surfaces what is happening across disconnected platforms
• Automated remediation for known issue patterns without human intervention
• Escalation routing that identifies which issues require human judgment versus which are rule-based resolvable
• Compliance monitoring and audit trail generation across infrastructure events

The agentic layer is the key distinction from traditional IT automation tools. Sustain agents do not just trigger alerts. They investigate, cross-reference, determine root cause, and either resolve or escalate based on confidence level. That is the difference between a monitoring tool and an agent: one tells you something is wrong, the other tries to fix it.

Why This Matters Beyond Enterprise IT

Sustain is an enterprise product. But the problem it solves is not exclusive to enterprises.

Every business that runs multiple software tools has some version of this problem. Your CRM does not talk to your project management tool. Your accounting software does not read your client portal. Your support desk does not update your operations dashboard. The manual work that bridges these gaps is invisible in most businesses — it lives in someone’s daily routine as a habit, not in any process document.

When that person leaves, the bridge breaks. When the business scales, the bridge becomes a bottleneck. When you are trying to understand what is actually happening in your operations, the bridge is invisible to any reporting tool.

Agentic AI applied to this layer — not replacing your software stack, but intelligently bridging it — is one of the clearest near-term opportunities in business automation. Sustain’s launch today is the enterprise version. The SMB version is available right now through Make, n8n, and AI agent frameworks that can wire your existing tools together without replacing them.

Story Two: Claude Opus 4.8 Dynamic Workflows — 750,000 Lines of Code in 11 Days

What the Developer Actually Did

Developer Jarred Sumner used Claude Opus 4.8’s dynamic workflow capability to migrate the Bun runtime from the Zig programming language to Rust. The codebase covered 750,000 lines of code. The migration completed in 11 days, with a 99.8% test suite pass rate.

For context: a migration of this complexity, handled manually by a senior engineering team, would typically take three to six months. The test pass rate on a major language migration would often require weeks of additional debugging to reach 99.8%. Eleven days and 99.8% is not an incremental improvement. It is a different order of magnitude.

What Dynamic Workflows Actually Are

Claude Opus 4.8 introduced dynamic workflows in its May 28 update. The capability automatically generates orchestration scripts and deploys multiple sub-agents for complex tasks.

Ken Takao, Lead Systems Engineer at a company using the feature, described it precisely: “Dynamic workflows fill the gap between firing off a single sub-agent and building out a full agent team. Plan to implementation just flows, so we can trust longer runs without losing control.”

The ultracode setting, activated for the Bun migration, allows the system to tackle large-scale projects like codebase migrations or security audits by automatically spawning and coordinating the right number of specialised agents for each component of the task. The developer defines the goal. The dynamic workflow determines the execution architecture.

The Implication for Non-Technical Business Owners

The Bun migration story is a developer story. Its implication extends well beyond development.

Dynamic workflows represent a new category of automation capability: give the system a complex, multi-step goal and let it determine how to break that goal into executable sub-tasks, spawn the agents needed to handle each, and coordinate them to completion. That capability applies to any sufficiently complex workflow, not just code.

Content audits across large archives. Due diligence across document libraries. Customer data reconciliation across multiple CRM snapshots. Market research synthesis from hundreds of sources. Any workflow that currently requires a team of people working in parallel on different components of a larger task is a candidate for dynamic workflow automation as this capability matures.

What Both Stories Share: The Broken Middle

The connection between Sustain and dynamic workflows is not obvious on the surface. One is an enterprise IT operations product. The other is a developer tool capability. But they are solving the same problem at different layers of the stack.

The broken middle of most organisations is the space between their tools, their systems, and their teams. It is filled with manual work that nobody designed, that nobody owns, and that nobody can easily see or measure. It is where information gets lost, where tasks fall through the cracks, and where scale creates chaos rather than efficiency.

Traditional automation addressed the edges of this problem: automate the repetitive, rule-based tasks at the beginning and end of workflows. Agentic AI, in both the Sustain model and the dynamic workflow model, is addressing the middle: the coordination, the judgment calls, the cross-system visibility, and the multi-step execution that previously required human orchestration.

That is the shift. And it is happening today, not in some projected future.

The Pattern Across This Week’s AI News

Step back and look at what has been announced in the past week alongside today’s stories:

• Microsoft Copilot embedded into NHS England for 505,000 healthcare workers (OS-level agent control)
• JPMorgan reporting 4 hours saved per employee per day from internal AI tools (knowledge worker productivity)
• GitHub Copilot switching to usage-based billing as agentic sessions become the primary value driver (infrastructure economics)
• Anthropic filing for a near-trillion-dollar IPO as enterprise revenue reaches $47 billion run-rate (market validation)
• Alteryx Agent Studio converting existing business logic into deployable agents without IT teams (democratisation)
• Publicis Sapient Sustain launching agentic IT operations (bridging disconnected systems)
• Claude Opus 4.8 dynamic workflows migrating 750,000 lines of code in 11 days at 99.8% accuracy (task complexity ceiling rising)

These are not separate stories. They are different dimensions of the same story: AI agents are moving into the core of how organisations operate, at every layer of the stack, at every scale of business, in every sector.

The businesses that understand this as a single coherent shift — rather than a collection of separate technology announcements — are the ones positioning correctly for the next 24 months.

What to Do With This Information Right Now

Map Your Own Broken Middle

The most valuable exercise any business can do right now is map where the manual bridging work lives in their operations. Ask every team member: what do you do every day that involves copying information from one place to another, or sending a message to bridge a gap between two systems that should communicate automatically? The answers will identify your highest-value automation targets.

Start With the Highest-Cost Manual Bridge

Once you have the map, rank by time consumed and frequency. The workflow that takes the most total hours per week and happens most often is the right first target. Build the automated bridge there first, measure the time recovered, and use that as the model for the next one.

Think in Terms of Goals, Not Rules

The dynamic workflow capability in Claude Opus 4.8 is a signal about where AI automation is heading: you define the goal, the system determines the execution architecture. Start shifting how you think about automation away from “If X happens do Y” and toward “The goal is Z, what does it take to get there?” That mental model will serve you better as the tools continue to develop.

Do Not Wait for the Enterprise-Grade Version

Sustain is an enterprise product. But the problem it solves is available for you to address right now with Make, n8n, and AI agent frameworks that cost a fraction of enterprise pricing. The SMB version of agentic IT operations is building automations that monitor your key business metrics, flag anomalies, and route issues to the right person without anyone having to check a dashboard. That is buildable today.

The Bottom Line on June 11, 2026

The AI automation stories breaking today are not about new capabilities that are coming. They are about new applications of capabilities that already exist, being deployed into the broken middle of how organisations actually operate.

Sustain is fixing the gap between disconnected enterprise systems. Dynamic workflows are fixing the gap between complex multi-step goals and the agent architecture needed to achieve them. Both are symptoms of the same maturation: AI automation moving from the edges of workflows into the core.

For any business owner reading this: your broken middle exists. It is costing you more than you can see in any report. The tools to fix it are available right now, at every budget level. The window to build before your competitors is open but not unlimited.

Frequently Asked Questions

What is Sapient Sustain and who is it designed for?

Sapient Sustain is an agentic AI platform from Publicis Sapient designed for enterprise IT operations and managed services teams. It deploys AI agents to handle incident detection, automated triage, cross-system visibility, and remediation for IT environments running multiple disconnected legacy and modern systems. It is primarily targeted at enterprises with complex IT infrastructure, though the operational problem it solves — manual bridging between disconnected systems — exists in businesses of any size.

What are Claude Opus 4.8 dynamic workflows?

Dynamic workflows are a capability in Claude Opus 4.8 that automatically generate orchestration scripts and deploy multiple specialised sub-agents for complex tasks. Instead of requiring the user to design the agent architecture, the system determines how to break a complex goal into sub-tasks, spawns the appropriate agents, and coordinates them to completion. The ultracode setting activates this capability for large-scale technical tasks like codebase migrations, security audits, and large-scale refactoring.

What is the ‘broken middle’ and how do I identify it in my business?

The broken middle refers to the manual work that bridges gaps between disconnected systems, teams, and workflows in an organisation. It is usually invisible because it lives in individual habits rather than documented processes. To identify it, ask your team: what do you do every day that involves copying information from one place to another, or sending a message because two systems that should communicate automatically do not? High-frequency, time-consuming answers are your highest-priority automation targets.

Can a small business use dynamic workflow automation without a technical team?

The ultracode version of dynamic workflows requires developer-level engagement. However, the underlying principle — defining a goal and letting the AI system determine the execution architecture — is increasingly available through no-code platforms. Make and n8n both support multi-agent workflows that can be configured without code. The capability gap between enterprise and SMB agentic automation is closing, and the most practically accessible version for non-technical operators is available today.

About the Author: Hamza Baig is the founder of Hexona Systems, an AI automation agency serving clients across six continents, and creator of the AI Automation Institute, where over 40,000 entrepreneurs have learned to build and scale automation businesses. He has been featured in GHL Top 50, Yahoo Finance, and Brainz Magazine. Follow him at @hamza_automates.