By leveraging the power of AI, Autobound enables sales teams to craft highly personalized, engaging emails that drive meaningful conversations and boost reply rates.

However, the journey to creating the perfect AI-generated content was not without its challenges.

‍

Autobound enables sales professionals to create hyper-personalized, relevant, and engaging outreach at scale based on millions of insights like news, social media, shared experiences, company initiatives, and more. With seamless integrations into popular sales tools like LinkedIn, Outreach, Salesloft, and Gmail, Autobound empowers sellers to boost their reply rates, build stronger relationships, and ultimately close more deals.

‍

As a company at the forefront of AI-driven sales engagement, Autobound recognized that the quality of their AI-generated content would be a key differentiator in their market.

However, they quickly realized that even state-of-the-art language models like GPT-4 struggled with the complex tasks required to generate compelling, personalized sales emails.

Autobound needed a solution that would allow them to rapidly iterate on their prompts and fine-tune their models to produce better content. Their existing workflow, which involved custom Python queries to measure latency and manage prompt versioning in Google Sheets, was a nightmare. Iterating on prompts using OpenAI’s playground and Azure AI Studio was equally challenging.

That's when Autobound discovered Vellum through a ChatGPT conversation - so meta, right?🤖

We sat down with Daniel Weiner, the founder at Autobound, to learn about their journey from a slow and complex AI development process to achieving a remarkable 20x acceleration in their LLM iteration cycle.

‍

Today, Autobound leverages all of Vellum's powerful features to speed up their AI development processes.

Prompt Collaboration

But the Vellum's Prompt Sandbox has become Autobound's favorite feature. It allows them to easily test various prompts and models for their email generations.

This functionality has made prompt management a breeze, saving the team countless hours and headaches.

Improving Prompts Using Live Data

Now, it’s very easy for the Autobound team to manage their deployed prompts from Vellum Deployments.

The ability to rapidly test prompts on real-world scenarios using live data has been invaluable for Autobound. This level of testing was previously impossible with other sandbox environments, giving them a significant advantage in fine-tuning their AI content generation system.

Evaluate Prompts Across Diverse Test Cases

The team has been using Vellum Evaluations to run several experiments with their prompts, ensuring they're of high quality and have low latency. Thanks to these experiments, they could speed up and reduce latency per LLM feature. We'll cover this further in the next section.

‍

20X Faster LLM Development

Vellum has accelerated Autobound's end-to-end LLM iteration cycle by at least 20 times, enabling them to rapidly test and refine their prompts on real-world scenarios using live data.

This level of efficiency was previously unattainable with other sandbox environments they’ve tried.

"Vellum has been a game-changer for us. The speed at which we can now iterate and improve our AI-generated content is incredible. It's allowed us to stay ahead of the curve and deliver truly personalized, engaging experiences for our customers."

‍Daniel Weiner, Founder @ Autobound

Reduced Latency by 5x

By iterating on their approach using Vellum, their team reduced the latency of their email generation system by an impressive 4-5x, from 30 seconds down to just 6-7 seconds per email.

This dramatic improvement was achieved by experimenting with different models, prompts, and fine-tuned models within Vellum's Prompt and Workflow platform.

As Autobound continues to push the boundaries of AI-powered sales engagement, their partnership with Vellum has become an integral part of their success story.

‍

Vellum has enabled more than 150 companies to build complex AI chatbot logic, evaluate with hundreds of test case, and deploy production-grade AI apps with confidence. If you’re looking to develop a reliable AI features, we’re here to help you.

If you want to try Vellum, book a demo here or reach out to us at support@vellum.ai if you have any questions.

We’re excited to see what you and your team builds with Vellum next!

‍