{"id":93070,"date":"2026-04-15T05:00:00","date_gmt":"2026-04-15T12:00:00","guid":{"rendered":"https:\/\/www.salesforce.com\/?p=93070"},"modified":"2026-04-15T04:48:07","modified_gmt":"2026-04-15T11:48:07","slug":"orchestration-next-frontier-for-agents","status":"publish","type":"post","link":"https:\/\/www.salesforce.com\/news\/stories\/orchestration-next-frontier-for-agents\/","title":{"rendered":"Why Orchestration Is the Next (Wildly Necessary) Frontier for Agents"},"content":{"rendered":"\n

Alcon, a global supplier of vision-aid products and technology for patients and ophthalmologists, is intent on becoming an Agentic Enterprise. And so, over the past year, various departments within Alcon developed hundreds of AI agents. <\/p>\n\n\n\n

But like other organizations on the agentic journey, Alcon soon realized that a collection of specialized AI agents presented its own challenges. \u201cEveryone went and built their own agents \u2014 business users and IT users alike,\u201d recalled Sreenivasa Patibandla, the company\u2019s Director of System Integrations and API. \u201cWe ended up with 900-plus agents built in silos. It\u2019s a security risk, first and foremost.\u201d And for a regulated medical device company, the hundreds of disparate agents created compliance issues, among other complications.\u00a0\u00a0<\/p>\n\n\n\n

\n

We ended up with 900-plus agents built in silos. It\u2019s a security risk, first and foremost.<\/p>\nSreenivasa Patibandla, Director of System Integrations and API, Alcon<\/cite><\/blockquote>\n\n\n\n

These are the problems solved by orchestration \u2013 the next great challenge for the Agentic Enterprise. The technology moves organizations beyond the development, deployment, and adoption of individual agents toward harmonious collaboration.<\/p>\n\n\n\n

Consider a musical metaphor. It\u2019s one thing for an AI soloist to sing the answer to a customer\u2019s simple question. It\u2019s quite another to create an AI jazz ensemble with the structure and coordination to produce a sonorous group improvisation. <\/p>\n\n\n\n

That, in essence, is what agentic AI orchestration<\/a> is all about: A group of talented, improvising agentic AI soloists work collaboratively, each handing off to the next player without missing a beat. And all the while, an AI bandleader \u2014 a superagent<\/a> \u2014 keeps the whole arrangement in the right key, on rhythm, and adhering to the same basic structure.<\/p>\n\n\n\n

When AI agents are orchestrated, businesses can leverage their creativity while ensuring they work together productively. At Alcon, the company has begun using a Salesforce orchestration enabler, Agent Fabric, to bring collaborative capabilities, and governance, to all those agents. <\/p>\n\n\n\n

Many other Agentic Enterprises have also recognized the need for AI orchestration and are taking steps to implement it. Besides Alcon, these companies include Royal Bank of Canada and the technology consulting firm Diabsolut.<\/p>\n\n\n\n

That\u2019s why Salesforce has released a series of products dedicated to solving the orchestration problem. That includes Agent Fabric<\/a>, which provides an enterprise agent control plane with unified agent management and advanced orchestration capabilities.<\/p>\n\n\n\n

Andrew Comstock is GM of Salesforce MuleSoft<\/a>, \u00a0a Salesforce company, and the creator of Agent Fabric. Comstock said that today\u2019s AI orchestration challenge is similar to one that IT managers have long faced \u2013 getting their services effectively operating in concert and not antagonistically.\u00a0\u00a0<\/p>\n\n\n\n

Today\u2019s AI agents can be thought of as the next generation of software. But there\u2019s an extra wrinkle. Apps and their APIs are deterministic \u2014 the same input always yields the same output.

But AI agents are nondeterministic, reasoning and adapting as they carry out their tasks. \u201cAn API does the same thing every time,\u201d Comstock said. \u201cAgents don\u2019t.\u201d

And so, coordinating and governing the activities of these next-generation, semiautonomous, unpredictable AI agents has become the crucial challenge that orchestration is meant to solve.<\/p>\n\n\n\n

The Curse of the Monolith<\/strong><\/h2>\n\n\n\n

For Gary Lerhaupt, Salesforce VP of Product Architecture, once an organization\u2019s first agent is working well, challenges will arise as the enterprise attempts to scale that success by simply adding more instructions and capabilities to a single agent, forcing it to handle too many disparate chores.<\/p>\n\n\n\n

Lerhaupt referred to these as \u201cmonolithic\u201d agents \u2014 jacks-of-all-trades that are masters of none. Overstuffing a single agent spreads the AI\u2019s attention too thin, leading to logic errors or hallucinations.<\/p>\n\n\n\n

\u201cThe ever-expanding monolithic agent eventually degrades,\u201d Lerhaupt explained. \u201cIt\u2019s an anti-pattern,\u201d he said, using the industry epithet for a solution that looks attractive on the surface but ultimately creates technical trouble.<\/p>\n\n\n\n

Ultimately, Agentic Enterprises must recognize that AI should follow the same organizational patterns as the humans they support. Just as a company wouldn\u2019t hire one person to be the CFO, the head of engineering, and a customer service rep, organizations shouldn\u2019t expect one AI agent to handle every corporate task.<\/p>\n\n\n\n

Human specialists know when to reach out to one another via Slack, email, or a quick meeting. Likewise, AI agents must be equally communicative team players. Without this coordination, Lerhaupt said, AI descends into the opposite extreme of the monolith: a \u201cmaze\u201d of disparate, single-task agents that customers and users must navigate on their own.<\/p>\n\n\n\n

\u201cYou don\u2019t want to have to tell your customers, ‘You can find Agent A over here and Agent B over there,’\u201d Lerhaupt said. \u201cThat is not a great user experience.\u201d<\/p>\n\n\n\n

RBC\u2019s Advisor Assistant: Orchestration in a Fiduciary Setting<\/strong><\/h2>\n\n\n\n

A company that has deftly avoided the monolith-maze double jeopardy is Royal Bank of Canada (RBC).<\/p>\n\n\n\n

Lerhaupt\u2019s team collaborated with RBC\u2019s Peter Herzog, who is the Technical Lead for the bank\u2019s U.S. Wealth Advisory business. Herzog\u2019s technologists support the wealth advisors who work directly with individuals and households to maximize the value of their financial holdings.<\/p>\n\n\n\n

Early in 2025, the Herzog team introduced an AI agent<\/a> that was carefully tailored to help advisors prepare for customer meetings and compile follow-up notes. The agent cut the time previously devoted to those activities in half. RBC advisors were so \u201cblown away\u201d by this efficiency, Herzog said, that his staff quickly pivoted to developing an array of specialized agents to make advisors even more effective and better informed.<\/p>\n\n\n\n

\u201cIn a wealth-management space that\u2019s highly regulated, that\u2019s no easy task,\u201d Herzog said of his team\u2019s efforts. In particular, the highly personal information of each financial household must be strictly firewalled. This ensures an AI agent can\u2019t disclose sensitive data to anyone outside the household other than the advisor, who holds fiduciary responsibilities. Orchestration that stresses trust and governance was essential.<\/p>\n\n\n\n

By late summer 2025 \u2014 following a rigorous review by RBC\u2019s regulatory compliance and governance teams \u2014 more than a dozen AI agents were in place, including specialists for market insights and customer-portfolio analysis. The superagent that coordinates these specialized tools is the RBC Advisor Assistant. It serves as the single point of entry for users.<\/p>\n\n\n\n

\u201cTypically, our wealth advisors are not technically advanced,\u201d Herzog said. \u201cThey expect to be able to type in a question and get an answer. We want to provide that experience for them.\u201d<\/p>\n\n\n\n