Developers Max Brodeur-Urbas and Rahul Behal assume that AI has the potential to automate a lot of business-relevant duties, however that most of the AI-powered automation instruments in the marketplace at present are unreliable and dear. Part of the issue is that customers count on an excessive amount of of AI, Brodeur-Urbas advised TechCrunch — as an example, they assume that it may deal with extremely specialised, area of interest workloads the place precision issues.
“If customers ever need to use AI for enterprise functions, the know-how actually has to don’t have any margin for error,” Brodeur-Urbas mentioned. “Leaving particular workflows utterly as much as AI will not be reasonable. Users could be paying for [an AI] to spin its wheels performing the identical Google search time and again.”
Still, Brodeur-Urbas, an ex-Microsoft software program engineer, and Behal, beforehand a software program developer at Amazon Web Services, thought at present’s AI had promising narrower purposes. So they began fascinated by methods they might squeeze what Brodeur-Urbas referred to as “actual worth” out of AI tech.
These concepts turned a wrapper for the open supply app Auto-GPT, then a proof-of-concept, and ultimately a startup: Gumloop. Gumloop automates repetitive workflows with AI, aiming to streamline fundamental duties.
“We began the corporate in a bed room in Vancouver as a aspect undertaking,” Brodeur-Urbas mentioned. “We had been attempting to resolve a quite simple downside for a gaggle of nontechnical individuals in a Discord server, and it spiralled into one thing bigger than we might have ever imagined.”
Gumloop gives a workflow builder that integrates with third-party apps and instruments together with GitHub, Gmail, Outlook, and X. Users can drag modular elements onto a canvas to construct automations, or select from prebuilt pipelines for duties like producing every day inventory stories and summarizing paperwork.
Brodeur-Urbas claims that groups at Instacart and Rippling are utilizing Gumloop for numerous use instances.
“Today, hundreds of customers depend on Gumloop as a core device for his or her enterprise,” he mentioned. “Giving nontechnical individuals the instruments to resolve their very own issues with out counting on engineers is the place we discovered market pull.”
There’s no scarcity of workflow automation instruments on the market. Parabola, Tines, Induced AI, and Nanonets come to thoughts. And on the horizon are “agentic” instruments from OpenAI and others, which promise to automate extra advanced duties end-to-end.
To stay nimble, Gumloop plans to maintain its workforce fairly small. The firm is hiring, however Brodeur-Urbas mentioned that the plan is to cap headcount at 10 individuals.
“Using AI to code let us have the throughput of a 20-person workforce and outpace opponents,” he claimed. “Our plan is to be a 10-person, billion-dollar firm.”
As it prepares to relocate from Vancouver to San Francisco, Gumloop has closed a $17 million Series A spherical led by Nexus Venture Partners with participation from First Round Capital, Y Combinator, and angel buyers together with Instacart co-founder Max Mullen and Databricks co-founder and chief architect Reynold Xin. To date, Gumloop has raised $20 million in capital.
“We didn’t want the cash in any respect,” Brodeur-Urbas mentioned. “Raising cash isn’t the purpose — constructing a product individuals love is. This new enterprise capital will assist us construct and scale that product even quicker.”
