A deep learning scientist who previously sold his startup to Snap has raised $4.5 million in seed funding for his latest venture, eSelf, which aims to develop real-time, video-based conversational AI agents. The Israeli startup is coming out of stealth mode with a focus on verticals such as education, sales, financial services, real estate, and health and wellness.

eSelf's initial pitch is that its agents can respond to queries in under 1.7 seconds, making them faster and more lifelike than voice responses from other leading companies like OpenAI. The startup has launched a "no-code" creation studio that allows customers to build their own agents, which can operate common apps like Calendly, Salesforce, and others to schedule appointments, access collateral content, and reference data just as a human assistant might do.

eSelf claims to have already powered "millions" of real-time conversations in stealth mode, with current customers including Christie's real estate and Brazilian bank AGI. The seed funding round was led by Explorer Investments, a firm out of Portugal, with participation from Ridge Ventures and strategic angel investors such as Eyal Manor, former VP of engineering at YouTube and chief product and engineering officer at Twilio.

eSelf is co-founded by Dr. Alan Bekker, CEO, and Eylon Shoshan, CTO, who met while working on voice recognition, neural networks, and deep learning projects. Bekker's background is notable, having previously sold his startup Voca to Snap for $100 million in 2020. He then became the head of Conversational AI at Snap, where he led a team that built conversational AI tech, which eventually became the My AI chatbot.

However, Bekker left Snap before the launch of My AI, which received mixed responses. He then started thinking about the next generation of AI agents, envisioning a more interactive and humanlike product using video and voice responses. This led to the founding of eSelf, which is now poised to enter the increasingly crowded "agentic" opportunity.

The tech world is rapidly becoming populated with companies targeting this space, including H, Eleven Labs, Amazon, Anthropic, /dev/agents, LinkedIn, and OpenAI. These companies are working on delivering responses that are not just more accurate but also more believable due to their form – a 'human' with a 'voice' delivering it. Startups and established players like Salesforce are also emerging with agents for specific use cases like shopping, assisting enterprise sales and revenue teams, and design.

While it's early days for this technology, it will be worth watching to see which companies catch on with paying users and which startups land investors to help them grow. One significant challenge in this space is the building and training of models, which can be a major cash drain. Bekker noted that eSelf has not built its own large language models (LLMs) from the ground up, instead fine-tuning Meta-created Llama models, among others.

As the "agentic" opportunity continues to evolve, eSelf's focus on real-time, video-based conversational AI agents positions it as a key player in this emerging market. With its seed funding in place, the startup is well-equipped to tackle the challenges ahead and make a significant impact in the industry.