Gusto Co-founder AI Automates Small Business Operations

Eddie Kim, co-founder and Head of Technology at Gusto, discussed Gusto Co-founder, a new AI-powered product designed to automate small business operations. Gusto, a Y Combinator Winter 2012 alumnus, has surpassed $1 billion in annual revenue and serves over 500,000 small businesses in the US, with one in five new businesses becoming Gusto customers.

What is Gusto Co-founder?

Gusto Co-founder aims to automate most business processes for small businesses. While many tasks are already handled by Gusto, the new product takes automation further. This includes:

• Payroll management: From start to finish.
• Time off approvals: Streamlining the process.
• Timesheet reminders: Prompting employees to submit on time.
• External integrations: For example, a tour guide business could receive a text if it's going to rain, and Gusto Co-founder could then email all customers to inform them and suggest bringing an umbrella.

Challenges in Designing AI for Small Businesses

Kim highlighted that many small businesses, and even most AI users, still treat AI as a "glorified search engine," asking questions and receiving replies. More advanced uses might involve research, report generation, or summarization, but these still largely involve a user-initiated query and a static response. The "agentic world" promised by AI has not yet materialized for the majority.

Gusto Co-founder aims to bridge this gap by moving beyond the "blank canvas problem" of open-ended AI tools. Instead of starting from scratch, it leverages existing Gusto functionalities in payroll, HR, time tracking, and scheduling. The product suggests ways to automate recurring tasks, such as running payroll (weekly for hourly employees, bi-monthly for salaried) or approving timesheets, often without the user needing to log into Gusto.

The Genesis of Gusto Co-founder

The initial idea for Gusto Co-founder stemmed from Kim's personal experience setting up an "open claw" (likely referring to an open-source AI agent or automation tool). He spent eight hours setting it up, including buying a Mac Mini and air-gapping it to prevent data deletion, a process only an engineer could easily manage. Despite the effort, he found himself using it as a glorified search engine. However, he was impressed by the ability to interact with the AI via Telegram, finding it a much better experience than using a web browser.

This experience highlighted a significant gap between reading about new technologies and actually experiencing them. Kim emphasized the importance for technical leaders, and even non-coders, to get hands-on with new tools to truly understand their potential.

The pivotal moment came during a layover in London, where a missed flight gave Kim five hours of uninterrupted time. Using "cloud code" (likely a code generation AI), he built a prototype. He realized the power of AI to build experiences based on intent, questioning why Gusto customers couldn't similarly prompt the system to build features for them.

The First Prototype and Its Evolution

The initial prototype was a chat prompt where customers could describe a desired user experience, such as:

• Sending a survey form to employees.
• Building a to-do tracking application.
• Creating a CRM.

The prototype would then generate a web application using Gusto's design system, making it appear as if Gusto had built it. This was essentially a CRUD (Create, Read, Update, Delete) web app builder that wrote and executed code, storing data brought in by the user.

However, this prototype didn't leverage Gusto's extensive data about its customers. The idea evolved from building web apps to creating automated workflows that utilize customer data. Gusto has aggregated data across industries (e.g., what a dentist might want to automate) and individual customer activity.

The core concept of Gusto Co-founder became centered around these automations, which run on a regular "heartbeat" or trigger. Kim noted that the technical design was heavily inspired by the open-source tools he used, particularly the cron job-like execution of LLMs. While open-source tools might use a less deterministic heartbeat, Gusto Co-founder offers scheduled jobs via regular cron for tasks requiring precise timing, such as payroll, and can detect which method is more suitable for a given request.

Selling AI to Small Businesses

Kim explained that selling Gusto Co-founder to small businesses is relatively straightforward because it addresses their immediate pain points: recurring, time-consuming tasks. He cited an example of a massage spa business that manually exports data from MindBody, manipulates it in Google Sheets for commission and tip calculations, and then inputs it into Gusto for payroll. This "work before the work" takes significant time each week.

When presented with the idea of Gusto Co-founder automating this process and providing a summary for final approval via SMS or Slack, small business owners "instantly get it." They recognize the value in saving an hour each week, allowing them to focus on growing their business. Unlike enterprise settings where AI might face resistance due to job displacement concerns, small business owners see it as an "all upside" opportunity to "do more with less."

The Future of Gusto Co-founder

Kim envisions Gusto Co-founder evolving beyond a business process automator to become a proactive "co-founder." This includes:

• Proactive advice: Notifying customers about compliance tasks or potential tax credits. For example, Gusto Co-founder identified a $50,000 R&D tax credit for a company called Cabana Pools, which they were unaware of.
• Market intelligence: Providing insights like competitor analysis or optimization suggestions for eBay sellers.
• Expanding beyond existing Gusto functions: Customers are already creating automations unrelated to Gusto's core services, such as a tour business receiving rain alerts and emailing customers.

Gusto recently launched Co-founder to 500 customers, following a pre-launch with a small group of 20 customers. The feedback has been overwhelmingly positive, particularly regarding the ability to manage operations via text message, mirroring Kim's own experience with Telegram.

Roadmap for Gusto Co-founder

Future plans include:

• More communication channels: Adding Telegram and WhatsApp, as SMS character limits can be restrictive.
• More connectors: Expanding integrations beyond QuickBooks, Notion, and Google Workspace to include vertical-specific tools like Curve Dental for dental practices.
• Broader accessibility: Opening the product to individuals who don't yet have a company or EIN, allowing them to automate side hustles or explore business ideas. Gusto Co-founder could then assist with formalizing the business, such as obtaining an EIN or registering as an employer.

The AI-Driven Development Process

Kim highlighted that the development of Gusto Co-founder itself was a testament to the power of AI. The prototype, initially built by Kim, organically attracted a team of five "AI builders" (four engineers and one designer). Over 10 weeks, they went from whiteboarding to a full-blown Tier 1 launch.

Key aspects of this accelerated development included:

• Minimal overhead: No meetings, text specs, or Figma designs. The goal was to avoid documentation.
• Constant collaboration: A 24/7 "permazoom" meeting allowed team members to hop in and out for discussions.
• Cloud code: Extensive use of AI for code generation.
• Rapid iteration: Ideas were immediately translated into pull requests. Code was frequently discarded if it didn't fit, a process deemed faster than traditional documentation and approval cycles.
• Blurred roles: The designer contributed production-grade code, and engineers were comfortable with design. AI tools allowed for quick prototyping and refinement, merging the crafts of design and engineering.

Kim advises other companies to embrace this approach, focusing on what not to do (e.g., excessive meetings, documentation) to achieve product velocity. While discipline in product vision remains crucial, AI enables rapid experimentation and tangible implementations, leading to better decisions than relying solely on abstract discussions or user research.

The Impact of AI on Small Business Creation

Kim believes AI marks a "step function" in making it easier to start and run a small business. While Gusto has already simplified back-office tasks like payroll, AI can now automate even more, including compliance, benefits setup, and HR rules. This frees business owners to focus on product development, customer acquisition, and growth. Census data already shows an increase in business applications, indicating a growing trend of people pursuing entrepreneurship.

Kim concluded by expressing his appreciation for small businesses, noting their unique and quirky contributions to communities.