How Lusha Deployed AI Agents

This is how a B2B tech company optimised for GenAI and successfully rolled it out.

How Lusha Deployed AI Agents

I sat down with Diana Atias for a conversation about her journey of innovation at Lusha and was in awe of her level of organization, her foresight, and her commitment to efficiency. I had never seen anything like it and became resolved to put pen to paper to share my observations.

First, some background:

Lusha is a business intelligence platform headquartered in the USA. Diana is a Senior Legal Counsel at the organization. She focuses on optimizing the legal team’s workflow and efficiency. At the start of last year, Diana set out to find a solution to address the endless stream of repetitive questions received by the legal team. These requests flowed at an unsustainable level, absorbing the team’s energy. Seeing the opportunity in the challenge, Diana’s instinct was to develop a Frequently Asked Questions (FAQ) document to address topics that the wider business frequently came to the legal team for support with.

This initial FAQ helped, but it wasn't enough. Diana realized the language used in these documents needed to be adapted for the sales teams. The legal terminology was too formal and not easily translatable into the sales context. After several iterations to make the language more accessible, the sales team still found it cumbersome to find answers in the FAQs, and they continued to reach out to the legal team directly.

To address this, Diana created a Slack channel for legal inquiries, hoping this would streamline the process. However, this still resulted in hundreds of repeat queries. Recognizing the need for a more robust solution, Diana attempted to create a chatbot on her own but found it inadequate. It was then that Diana set out on her search for a more sophisticated solution and was recommended Flank.

In collaboration with Flank’s team, Diana implemented AI agents designed specifically for their needs. These AI agents could learn the contents of the FAQ documents and use legal resources for research. The onboarding process was rapid, smooth, and intuitive. This quick and painless onboarding was a significant factor in its successful adoption.

With the AI agents in place, Diana and her team conducted internal stress tests. They invited various team members to ask questions, both common and complex, to ensure the AI agents could handle a range of inquiries. The feedback was positive, and after some fine-tuning, they rolled it out to the entire go-to-market team.

One of the critical challenges was ensuring the AI agents could provide precise and concise answers. The legal answers were often lengthy and complex, leading users to double-check with the legal team for clarity. Diana worked closely with the development team to refine the responses, making sure they were as straightforward as possible without losing the necessary legal nuance.

Another challenge was fostering adoption among the sales and other teams. Diana found that while some were hesitant at first, the convenience and efficiency of the AI agents soon won them over. By positioning the AI agents as a tool in beta and encouraging feedback, she was able to gradually build trust and reliance on the system.

Diana’s efforts have not only optimized the legal team’s workflow but also set a precedent for other departments looking to streamline their processes. Her story is a testament to the power of proactive problem-solving and the effective use of technology to enhance organizational efficiency.

It’s safe to say that Diana’s journey with Flank’s AI agents shows the importance of adaptability, user-centric design, and continuous improvement. Her strategic approach and dedication to efficiency have significantly impacted the legal team’s productivity, providing a valuable blueprint for others to follow.

Want to find out about our AI agents? See here.