Reinventing Your Business for the Age of AI Agents

The rise of AI agents is more than just a technological advancement; it’s a fundamental shift in how businesses can operate, and for the first time, scale is no longer an inherent defense. Startups, with their agility and a fresh “agentic-first” perspective, are uniquely positioned to disrupt established incumbents. By designing their core business processes to leverage AI, even small teams can achieve remarkable feats previously thought impossible for their size.

In recent interviews with Aaron Levie, CEO of Box.com, shared his view that simply agentifying existing processes will drive some benefits, but the real game changer is reinventing business processes that couldn’t have been done by humans as quickly and accurately.  Here is one video that I’d encourage you to watch in full. It echoes many of the things that we’ve been saying about AI and agents over the last couple of years. To get validation from someone like Aaron is reassuring.

Aaron Levie, CEO, Box.com

“The result isnʼt just doing more with less. Itʼs achieving what was previously unimaginable.” 

The Dawn of Agentic-First Businesses

Consider the success of SaaStr, a community for SaaS executives and entrepreneurs, founded by Jason Lemkin. He built a business generating $25 million in annual recurring revenue (ARR) with a lean team of just five employees, largely by redesigning every aspect of his operations around AI. From analyzing thousands of conference presentation submissions and building intricate schedules to crafting session blurbs, promotional artwork, and speaker bios, AI handles it all. The result? Faster, cheaper, and more accurate outcomes than what traditional outsourced teams could deliver. This entrepreneurial spirit, akin to a “digital gold rush,” is now making it easier to develop and market tech solutions, fostering a new wave of innovation.

Box.com have put out a White Paper called “Becoming an AI-First Company” DOWNLOAD  LINK  

The key takeaways are:

• Being AI-first means redesigning workflows, not just automating tasks
• AI frees people up to focus on innovation by handling routine work
• Strong data management unlocks valuable AI insights
• Success requires proper governance and change management

Why Agentic Processes Aren’t Just “People Processes” with AI Sprinkled On Top

The temptation to simply overlay AI agents onto existing human-designed business processes is strong. While this might offer some cost savings and efficiency gains—what some call “your mess for less” —it fundamentally misses the point.

Traditional processes were crafted with human limitations in mind:

Limited Information Processing

Humans can’t realistically absorb and synthesize 200 pages of background material or glean insights from four disparate systems simultaneously.

Repetitive Task Fatigue

Unlike AI, people get bored with repetitive tasks and are unlikely to explore a problem 100 different ways to find the absolute best solution without getting bored.

Information Lag

Humans may not always have the most up-to-date information on policies or pricing.

Inherent Biases

Emotions, personal relationships, or even a “bad hair day” can introduce natural biases into human decision-making.

The danger lies in accelerating a poor process; customers will gravitate towards organizations with great support agents that are easy to deal with. Similarly, employees desire workplaces where agents reduce “busy-work”.

Redesigning for AI: Leaning into Strengths

To truly harness AI’s power, you need to step back and redesign your business processes from an agentic perspective. This means optimizing for AI’s strengths: its massive analytical capabilities, tireless processing, and unbiased execution. As Jason Lemkin notes, he feeds all his writing and presentations into an AI engine not because it’s inherently “better” than him, but because it “remembers everything.” This re-evaluation of business processes, where you can completely rethink your operations and design exceptional experiences, is a significant opportunity.

As Aaron notes in the white paper, “Weʼve identified that the companies most successful at adopting AI tend to follow certain key principles. These principles extend beyond mere automation, envisioning AI as a transformative force.”

Imagine the possibilities across various departments:

Customer Success

AI can analyze vast customer interaction data to proactively identify potential issues, suggest personalized solutions, and even anticipate future needs.

HR

AI can streamline onboarding, answer employee policy questions instantly based on employee location, and manage complex requests like paid time off (PTO) based on intricate rules across different locations.

Legal

Agents can sift through mountains of legal documents, identify relevant clauses, and even draft initial responses, accelerating legal processes significantly.

Marketing

AI can analyze market trends, personalize customer communications at scale, and even optimize campaign performance by identifying the most effective messaging.

Not Every Problem is an Agent Problem

In the enthusiasm to “agentify” everything, it’s crucial to remember that not every business process is best served by a pure AI agent. Most agents are a combination of AI and workflows, highlighting the importance of finding the right balance. Some processes are simply better, faster, simpler, and easier for the user as a workflow (e.g., a button that opens a form and a workflow that processes it). For instance, while we successfully “agentified” the HR process for booking PTO as a learning exercise, we found that a simple form with a workflow was a more efficient and user-friendly solution for that particular task.

Here are some criteria to help determine if an AI agent is the right fit:

Understanding User Language

If your users struggle with precise terminology, AI agents can bridge that gap by interpreting natural, varied language and translating it into actionable inputs for your system. This means users can express themselves comfortably, and the agent will translate their meaning into the format your system needs.

Working with Unstructured Data

For tasks requiring the digestion of free-form text, conversations, or messy data, AI agents are great at processing messy data to find important insights. This makes your workflows smoother by automatically understanding complex information that isn’t neatly organized.

Automating Tricky Logic

Do you have complex rules or processes that are hard to program traditionally? AI agents can perform advanced reasoning that would be very difficult to code.

Managing Complex Validation

Do your forms or processes have complicated rules that depend on many different pieces of information? AI agents can handle these tricky checks easily. Just tell the agent what the final result should look like and how to get there, and it will manage the detailed validations, making development and upkeep much simpler than writing lots of conditional code.

Assisting with Flexible Planning

Are there situations that need strategic thinking, negotiation, or adaptable planning that can’t be put into fixed rules? AI agents can help with dynamic planning. For example, in sales, if a customer wants a certain number of licenses but has a strict budget, an agent can help figure out the best deal by looking at things like license count, budget limits, and different contract lengths (like one-, two-, or three-year terms). This involves flexible decision-making beyond simple automation.

Improving Data Accuracy and User Experience

Is getting accurate data crucial, and do users sometimes make mistakes or skip fields? AI agents can significantly boost data quality and make things easier for users. They can guide users to the right answers, even if their initial input is unclear. Plus, agents can reduce manual effort by using context to intelligently pre-fill information, making the process smoother and less error-prone. Organizations with strong data governance, where data quality is paramount, are well-positioned to maximize the benefits of agents.

Building Your Organization’s Operational Graph for the AI Era

To truly embrace this agentic future, organizations need a precise understanding of their current state—an “operational graph” of their business operations. This encompasses not just high-level processes but also the underlying data structures and application metadata and dependencies that dictate how work actually gets done. Many companies operate in a “black box,” unable to visualize what they’ve built, how it connects, or if it’s even correct. This lack of visibility stifles agility and escalates project risks.

To be clear, building the operational graph is far more than simply using AI to interrogate the org metadata. What is missing is the relationships between the metadata. These dependencies are not easily analyzed because every metadata type stores references in different formats and structures. This is part of the reason that Salesforce’s work on the DependencyAPI was halted over 3 years ago. 

Instead, it has taken 100’s of FTE years of R&D effort by ISVs to calculate the dependencies. Many of the ISVs analyze where metadata is used, but the operational graph needs to know “how” and “where”. Only Elements analyzes to this level.

The Elements core capabilities of process mapping and the metadata dependency analysis have unlocked the ability to automatically generate the operational graph using a feature called Configuration Mining. This innovative tool can automatically generate documentation from metadata, including process diagrams, data models, and order of execution diagrams. This is the “holy grail” of business process management and metadata management that has been sought for decades.

Configuration Mining provides an on-demand “blueprint” to visualize, understand, and act upon your Salesforce configuration. It essentially reverse-engineers business processes directly from your Salesforce metadata, offering a current, accurate view of how your organization works.

Agents Finding Agents

Once this comprehensive operational graph is established, Elements can then leverage a feature like Agent Finder. This capability allows the system to analyze your documented processes and automatically identify potential use cases for AI agents. It goes beyond mere suggestions by providing reasoning and confidence levels for its recommendations.

This means you can pinpoint exactly where agents can have the greatest impact, optimizing processes before applying AI to them, and ensuring that your investment in AI delivers tangible, reliable results. It’s about changing the documentation game, moving from static, outdated reports to dynamic, on-demand insights that truly support business transformation

In a 90 minute workshop, a dozen Certified Technical Architects – the highest certified individuals in the ecosystem – identified 5 agent use cases from some process diagrams they were given. Our AI Agent Finder, in less than 2 minutes, identified an additional 3 agent use cases.

Final Word

The takeaway is clear: not every problem demands an AI agent, and not every agent needs to be 100% AI. However, every business problem should now be re-evaluated through an “agentic-first” lens, rather than being constrained by traditional human-centric process design.

This shift is akin to the “digital gold rush” we saw when it became easier to develop and market tech solutions, fostering an entrepreneurial spirit. Just as companies sought digital transformation and retooled employees with consumer-style applications, bringing B2B back into vogue, the “operational graph” of your business will be crucial for understanding how to leverage AI

Cover Photo by Suzanne D. Williams on Unsplash