These days, public discussions about technology are dominated by generative AI. Ema, a San Francisco-based firm, sees it as more than simply a trend. It came out of hiding today with a product that shares its name and aims to revolutionize the way we engage with artificial intelligence, particularly generative AI.
“Our goal is to build a universal AI employee,” stated Surojit Chatterjee, CEO and co-founder, during an interview. “We aim to automate routine operations that all businesses undergo so that workers can focus on higher-value, more strategic endeavors.”
Together with its investors, the company is backing its claims with actions and revenue: To disprove any claims of vaporware, it has already secured $25 million from a superb list of investors, including clients it secretly acquired while remaining under the radar, such as Envoy Global, TrueLayer, and Moneyview.
In terms of what Ema can accomplish, these companies are putting it to use in a variety of applications, from internal productivity tools for staff to customer service (which includes providing technical help to users and tracking, among other things). EmaFusion and the Generative Workflow Engine (GWE) are two of Ema’s solutions that may “emulate human responses” and improve with time and user input.
Chatterjee argues that it’s more than simply RPA(robotic process automation), AI for specific job acceleration, which is even more outdated, or yet another GenAI accuracy gaffe ripe for online mockery.
According to Chatterjee, Ema (an acronym for “enterprise machine assistant”) uses its own “smaller, domain-specific models” in conjunction with more than 30 large language models. This patent-pending platform aims to solve all the problems related to accuracy, hallucination, data protection, and more.
Many more names are being added to Ema’s cap table in this initial round. There are several venture capital firms involved, including Accel, Venture Highway, Section 32, AME Cloud Ventures, Prosus Ventures, Wipro Ventures, Maum Group, Frontier Ventures, and Firebolt Ventures. Moreover, some prominent individuals have provided financial support, including Sheryl Sandberg, Divesh Makan, Dustin Moskovitz, Jerry Yang, and David Baszucki.
Numerous businesses are currently developing GenAI tools for organizations. Some focus on certain verticals or use cases, while others are taking a more ambitious approach, similar to Ema’s home-run style swings. You may be asking why these investors are showing interest in this specific GenAI venture, and part of the answer may lie in the fact that they are currently generating revenue. However, this is due in part to the team’s history as well.
Before joining Ema, Chatterjee served as Coinbase’s chief product officer in the phase preceding the company’s initial public offering. Before that, he oversaw Google’s mobile advertising and commerce divisions as VP of product. Machine learning, adtech, and corporate software are just a few of the fields in which he has filed for over 40 patents.
Similarly noteworthy is the experience of Souvik Sen, another co-founder and head of engineering at Ema. He most recently served as VP of engineering at Okta, where he was responsible for data, machine learning, and devices. Before that, he was the engineering lead for data and machine learning at Google, where he oversaw privacy and safety initiatives. He has 37 patents to his name.
Both of their backgrounds give validity to the company’s goals and the possibility that it can achieve them. On the other hand, it reveals an array of information that could very well affect its development.
Take Chatterjee’s knowledge of e-commerce and adtech as an example. It seems to reason that they will play a role in how Ema could develop if it takes off since they form the bedrock of how many companies engage with customers today.
However, the firm may have a higher chance of avoiding data protection and privacy issues if its creator has prior experience with these areas. We can only hope, though! Artificial intelligence (AI) is at the heart of this Silicon Valley venture, which will concentrate on the task and how to leverage technology to do it.
At the time, it is noteworthy to observe ambitious companies striving to develop products that integrate several LLM silos to attain more sophisticated outcomes. It might be an indication that LLMs are becoming more commoditized and interchangeable than you would think.
Furthermore, investors believe the startup’s potential diversification—the ability to cut across diverse use cases—could help it grow its business and be more beneficial overall.
Ashutosh Sharma, head of investments for Prosus Ventures in India, told TechCrunch that while most point GenAI solutions are valuable for specific use cases, they aren’t always easy to expand across or even to adjacent use cases. Moreover, big enterprises are concerned about data fragmentation and the number of applications that could access their sensitive information. “Ema can provide the best return on investment while solving these issues and delivering high accuracy.”
