How Gumloop Enables Anyone to Make AI Workflows

"I don’t think about Reducto at all—in the best way possible. It just works. It’s perfect." - Max Brodeur-Urbas, Founder of Gumloop

In modern enterprise, the appetite for efficiency automations is enormous—but building custom AI workflows still feels out of reach for most employees. For non-technical teams buried in manual work, the ability to create AI-powered tools on their own terms is a game-changer.

Max Brodeur-Urbas, the founder of Gumloop, saw this first-hand when he released a side project AutoGPT, an open-sourced UI for an agent framework that allowed people to have AI just handle things for them. He saw a hunger from early non-technical users who wanted to utilize AI, but had plenty of questions about how to use AutoGPT.

“Almost everything they were asking could’ve been solved with a Python script or a couple of API calls, but they couldn’t get themselves there because coding was just a nonstarter. So I made a reusable framework” said Max.

That’s when Gumloop was born: an AI workflow builder with an intuitive drag-and-drop interface that makes it easy for anyone to automate repetitive work. Teams at Instacart and Webflow use it to rapidly scale AI adoption across the org—saving time and unlocking better results.

Invisible Document Infrastructure that Just Works

Early on, Gumloop offered basic support for reading PDFs. But as usage grew, so did expectations—and customers were running into more and more document edge cases.

“If you pass garbage into the AI model, you're gonna get garbage out,” Max explained. “We needed a better way to get information out of PDFs.”

Gumloop knew it didn’t make sense to build their own parsing engine. Their product philosophy was to move fast, integrate the best tools, and focus on delivering value through the interface.

Today, Reducto powers Gumloop’s “advanced PDF reading” node option, enabling high-fidelity document ingestion in any user-generated workflow. This lets users pull data from a slew of sources like Google Drive, Sheets, Gmail attachments, and much more.

"Our customers just think Gumloop is amazing at parsing PDFs... we don’t need to worry about Reducto going down or anything. It works perfectly and our customers are happy" said Max.

How Teams Are Using Documents in Surprising Ways

What makes Gumloop so compelling isn’t just the interface—it’s what people build with it. Max regularly hosts office hours during large rollouts, watching firsthand how nontechnical employees create automations that solve real bottlenecks.

One team built a workflow that scans GitHub issues for release blockers—flagging anything marked “shipping this week” that wasn’t going to be finished in time. Previously, a team lead did this manually every morning. Now it runs in the background.

Other creative workflow examples include:

• Compiling unread Slack messages in a specific channel with no reactions from today
• Finding all the churned customers in Stripe from the last 6 months that didn’t spend over $1k
• Parsing the receipts in a zipped file and calculating monthly spend
• Detecting mentions of our company across all social media channels, and flagging important instances

Even engineering teams are embracing Gumloop’s no-code approach. An internal audit of Instacart’s usage showed a surprisingly even distribution of workflow creators across departments—including nontechnical and technical roles alike.

“The power of Gumloop is letting people build their own tools that don’t exist, because it’s too niche and too specific for any vertical SaaS to tackle” said Max.

Example role distribution of Gumloop users from Instacart case

A Shared Philosophy: Speed and Shipping

Gumloop is built by a tiny, fast-moving team—just 8 people, with only 4 engineers. When breaking into enterprise accounts, Gumloop stood out by shipping faster than anyone else—without compromising quality. That kind of velocity required working with partners who shared the same philosophy.

“We thought the Reducto team was the only team that shipped as fast as we do. So we wanted to go with the best software.”

That shared emphasis on speed and reliability has made the integration seamless—and lets Gumloop stay focused on delivering delightful user experiences and the best quality automations for its customers.

What’s Next: Zero Learning Curve AI Adoption

Gumloop’s roadmap centers on making automation feel effortless—like Cursor for AI workflows. Their upcoming releases, like Gummie AI and MCP Nodes, are designed to remove friction entirely. Users can describe what they want, and the system builds the automation—end to end. Users can just use natural language to describe what they’re looking for, and Gumloop will automatically generate code for you to use, knowing exactly which nodes and MCPs to chain together.

“We try to build a solution that will let them address their own problems. We want to have a zero learning curve.”

By designing a system that’s both dynamic and supportive, Gumloop unlocks a new level of creativity—allowing anyone to experiment, iterate, and build powerful AI workflows on their own terms. With the information needed oftentimes stored in documents or other forms of unstructured data, Gumloop partnered with Reducto to make that data instantly usable. From parsing PDFs to extracting key details with precision, this integration ensures that users can connect the dots faster—turning messy inputs into structured building blocks for automation.

The future of work is AI-native, and Gumloop and Reducto are helping teams build it, one workflow at a time.