Max Brodeur-Urbas and Rahul Behal, developers in the tech industry, recognise the potential of AI to automate numerous business tasks. However, they express concern that many current automation tools driven by AI are inconsistent and expensive. Brodeur-Urbas points out that users often have unrealistic expectations of AI, believing that it can effortlessly manage highly specialised tasks that require precision.
“For AI to be deployed effectively in enterprise scenarios, it must operate with zero margin for error,” Brodeur-Urbas stated. “Relying entirely on AI for certain workflows is impractical. Users would essentially be paying for an AI to repeatedly perform the same search results on Google.”
Despite this, Brodeur-Urbas, a former software engineer at Microsoft, alongside Behal who previously worked at Amazon Web Services, identified certain focused applications of AI that show promise. They began exploring how to extract what Brodeur-Urbas termed “real value” from AI technologies.
Their ideas evolved into a wrapper for the open-source application Auto-GPT, leading to a proof-of-concept, and ultimately to the creation of a startup: Gumloop. Gumloop aims to utilise AI to automate repetitive workflows, enhancing efficiency in basic tasks.
“We initiated the company as a side project from a bedroom in Vancouver,” Brodeur-Urbas shared. “Our goal was to address a simple issue for a group of nontechnical individuals on a Discord server, and the concept expanded far beyond our initial expectations.”
Gumloop features a workflow builder that seamlessly integrates with various third-party applications, including GitHub, Gmail, Outlook, and X. Users can effortlessly drag and drop modular components onto a canvas to create automations or select from pre-configured pipelines for tasks like generating daily stock reports and summarising documents.
Brodeur-Urbas notes that teams within companies like Instacart and Rippling are currently leveraging Gumloop for various applications.
“At present, thousands of users depend on Gumloop as an essential tool for their business,” he commented. “Empowering nontechnical users to resolve their challenges independently without needing engineers is where we’ve identified significant market demand.”
The market is saturated with workflow automation tools such as Parabola, Tines, Induced AI, and Nanonets. Additionally, advanced “agentic” tools from OpenAI and others are emerging, promising to automate intricate tasks from start to finish.
To maintain agility, Gumloop intends to keep its workforce deliberately small. While hiring is ongoing, Brodeur-Urbas mentioned that they plan to limit the number of employees to 10.
“Utilising AI for coding allows us to achieve the output equivalent to a 20-person team, thus outpacing our competition,” he asserted. “Our vision is to become a 10-person, billion-dollar enterprise.”
As the company prepares to move its operations from Vancouver to San Francisco, Gumloop has secured a $17 million Series A funding round led by Nexus Venture Partners, with contributions from First Round Capital, Y Combinator, and angel investors including Instacart co-founder Max Mullen and Databricks co-founder and chief architect Reynold Xin. So far, Gumloop has raised a total of $20 million in funding.
“We didn’t actually need the funding,” Brodeur-Urbas explained. “Our primary aim isn’t just to raise capital — it is to develop a product that users love. This new venture capital will accelerate our ability to build and expand that product.”