Sybill’s AI reads customer behavior to automate sales workflows

Enterprises want to accelerate their sales, but more often than not, the reps responsible for talking to customers and converting them are bogged down with tedious administrative tasks. Today, Sybill, a startup solving this problem with a novel AI assistant, announced it has raised $11 million in Series A funding from Greycroft, Neotribe Ventures, Powerhouse Ventures and Uncorrelated Ventures.

While there are plenty of vendors that operate in the coveted “conversation intelligence” space, allowing account executives to automate tasks and better connect with their prospects, Sybill claims to have solved the problem of behavioral understanding. Essentially, the company’s AI copilot can read the intent and behavior patterns of potential customers, giving sellers detailed insights into what they are interested in and automating several tasks to help them close the deal.

The startup, founded by AI experts from Stanford, UC San Diego and Harvard, already claims to have roped in more than 500 enterprises as customers. It will use the capital from this round to grow its footprint, capture buyer insights on a deeper level and automate even more tasks for sellers. 

Capturing insights from sales conversations

Currently, when a sales rep handles a customer account, they manage a variety of different administrative tasks throughout the deal lifecycle, right from transcribing and analyzing the conversations with the prospect to extracting relevant information and feeding it into different platforms, including CRMs like Salesforce and Hubspot.

While the process seems pretty straightforward, the reality is that these tasks take a lot of time, keeping sales teams from selling the product or service. According to Salesforce’s recent State of Sales survey, reps spend just 28% of their week actually selling and the majority of their time goes to tasks like internal meetings, manual data entry, research and planning.

Former Stanford lecturer Gorish Aggarwal first saw this challenge when trying to handle multiple tasks while teaching remote classes. Subsequently, he and fellow AI experts Nishit Asnani, Soumyarka Mondal and Mehak Aggarwal conducted extensive research to understand the various tasks that get in the way of sellers doing their job. This led them to come up with Sybill in 2020.

At the core, Sybill’s AI copilot sits in the meeting when a sales rep goes on a call with a potential customer and records the whole discussion. Once the call ends, it produces a detailed report of the conversation, complete with a summary and transcription. This sounds pretty similar to Otter, but Sybill goes a step ahead by providing detailed insights into the behavior of the parties who joined the call — while keeping the conversation in context.

“We’ve built our own computer vision models to understand user behavior in the context of the conversation. These detect non-verbal cues like smiles, nods, leaning in/out, distraction, etc. and match it up with the text to get a deeper understanding of points of interest and pain points,” Aggarwal told VentureBeat.

The CEO says sellers can hover over a dedicated bar under the call recording to see how their prospect interacted and reacted throughout the conversation. They can easily figure out what part of the conversation kept them engaged and excited as well as where they lost attention. The generated call summary also features a list of pain points and interests of prospects, based on the conversation, to help users understand what they need to do next to quickly close the deal.

While the CEO did not share the names of the models powering this sales-centric experience, he did note that it all happens with a combination of internal and external models, including large language models, entity detection models and other natural language models.

Insights power automated actions

Once the call insights are captured and generated, Sybill’s AI assistant combines it with other information from the same account, including past calls and emails and the stage of the deal, and uses everything to draft a personalized follow-up email. It even learns from past emails to help the user quickly tweak the writing style and tone of the message according to their need.

From here, the seller can instantly send the AI-generated email, without spending time to manually craft it for the customer in question. They can also see all the insights and context captured by the tool across conversations and emails, complete with relevant elements like the budget, authority, need and timeline and integrate. This information gives a bird’s eye view of the account and the stage it is in and is automatically pushed into the CRM connected by the seller. 

This saves the seller from the hassle of manually entering relevant information about the opportunity into the CRM and updating it from time to time. As the calls and email interactions continue, the information in the CRM automatically gets updated. 

Beyond this, the insights captured by Sybill AI are also pushed into tools like Slack, allowing sellers to keep relevant parties, such as sales managers, informed about how the deal is progressing.

 Sybill saves five to ten hours every week

Since launching the company, Sybill has seen significant growth and has roped in over 500 small businesses and enterprises as customers, including Spot, SupportLogic, Zenhub and Amper. Over the last 18 months, the company says its revenue grew 15-fold, with over 60% of new revenue coming from customer referrals.

The company charges on a per-seat basis and has two plans on offer: one for those looking to solve basic call notetaking needs and the other for large players requiring an AI sales assistant, complete with follow-up emails, deal summaries, CRM sync and other features. It says the advanced plan of the tool is already saving sales reps five to ten hours every week —  equalling $20K-$40K annually, based on typical account executive compensation.  

With the fresh funding, which takes Sybill’s total capital to nearly $15 million, the company will focus on further expanding the capabilities of the AI assistant and automating more administrative tasks to help reps close more deals. 

“We will invest in building our brand as the go-to place to learn about AI in sales. Sybill is building the AI sales assistant, and our audience is hungry to learn about how they can best leverage AI to up-level themselves and their teams. Sales motion is not benefiting from the current AI revolution as much as engineering or marketing is, and we will change that,” the founder added.

Other notable players handling sales-specific conversational intelligence are Gong, Echo AI (formerly Pathlight) and Zoominfo-acquired Chorus. However, the CEO was quick to point out that most of these offerings are solutions that provide insights for decision-making, not for taking automated actions like Sybill.