How AiSDR Saved 5,000+ Hours with Gumloop

AiSDR, an AI-powered sales development platform, helps sales teams automate inbound and outbound sales processes: everything from prospecting and personalized outreach to lead engagement and meeting scheduling.

Naturally, as an AI-powered platform, AiSDR understands the value of building AI-powered workflows. They wanted to find an AI automation platform for their internal use, but weren’t interested in building or managing additional backend infrastructure to make that happen.

“We were not interested in building out the infrastructure for agents,” says Oleg Zaremba, the co-founder and CTO of AiSDR. “The challenge that we faced was wanting to experiment with AI quickly, and to build a production-ready solution that can be triggered by non-technical users.”

Why AiSDR Chose Gumloop

Gumloop came onto Zaremba’s radar when he saw a post on X from Tobi Lütke, the CEO of Shopify, about how Gumloop was being rolled out for Shopify across the whole organization. After evaluating Gumloop, AiSDR found it checked all their boxes. In particular, AiSDR uses a combination of different large language models — from Google, Anthropic, and OpenAI — and Gumloop’s model-agnostic approach allowed AiSDR to use all of them seamlessly.

“The benefit of using Gumloop instead of building it all on our own is that it’s much faster for us and we have a lot more ability to experiment,” Zaremba says.

How AiSDR Uses Gumloop

AiSDR has already implemented dozens of different workflows, and they’re just getting started. Here are some of the most impactful ways they’re currently using Gumloop.

Feature announcement workflow

AiSDR is constantly releasing new features, and with Gumloop, they’re able to orchestrate a complex workflow that allows them to enable new features and release them to customers.

“When we announce new features, we want to be able to automatically enable them for users, we want to be able to notify those users, and then also record the information about the fact that we have enabled that for the users,” says Zaremba.

To do this, AiSDR integrates PostHog, which then makes a cURL API call to Gumloop. From there, they fetch all the necessary data, process the information sent from Gumloop through a model, and craft a Slack message. That Slack message gets sent to a group chat with the customer, and then another API request gets sent to enable the feature.

Building ICPs

One of AiSDR’s most important workflows involves building out ideal customer profiles (ICPs) for their customers. AiSDR’s customers provide several inputs — including website URLs, internal documentation in the form of PDFs and doc files, examples of brand messaging. AiSDR then runs this information through a multi-step Gumloop workflow that analyzes this information to return a very detailed ICP. Customers can then use this ICP inside the AiSDR platform to optimize their messaging.

The AI Strategist

AiSDR has also combined Gumloop and their own APIs to implement what they call their “AI strategist,” which allows customers to use AiSDR’s features — “without actually needing to know the product.”

“This is basically our way to make all of the plays, all of the functionality of AiSDR available through natural language, and through a one-click interface,” says Zaremba.

To use the AI strategist, customers input their website and other additional information. The AI strategist then uses a Gumloop workflow to find the target audience, as well as optimal messaging, filters, and campaign types. It also suggests 5–10 fully customized campaigns that can then be launched with one click inside AiSDR.

Gumloop’s Impact, By the Numbers

• 12+: the number of Gumloop flows that AiSDR currently has in production
• 5,000+: hours of manual work saved by just one of AiSDR’s Gumloop flows

“We ran this Gumloop flow more than 2,000 times and saved 5,000 hours of manual work.” — Bohdan Kalymon, Customer Success Operations, AiSDR

AiSDR 🤝 Gumloop

With Gumloop as their AI automation platform, AiSDR has been able to rapidly iterate on dozens of different AI-powered workflows, without needing to build or manage their own backend infrastructure. They’ve also been able to deliver a more friendly, intuitive experience to their non-technical users.