In today’s digital-first world, customer service is undergoing a seismic shift, with businesses increasingly turning to AI-driven solutions to meet rising expectations for speed and efficiency. French startup Rounded is leading this charge, developing AI voice agents that promise to transform how small businesses and entrepreneurs interact with their customers. By automating routine tasks and delivering instant support, Rounded’s technology not only enhances operational efficiency but also redefines the customer experience. This article explores Rounded’s journey, its innovative platform, and the broader implications of AI voice agents for businesses navigating the digital landscape.
From Web3 to AI Voice Agents: Rounded’s Journey
Rounded began its journey in the web3 space but quickly pivoted to AI voice agents in June 2023. The initial idea was straightforward: combine ChatGPT with a speech-to-text transcriber and a text-to-speech synthesizer to create a voice agent. “The idea was that we’d simply put ChatGPT after a transcriber and before a synthesizer, and it would be valuable,” said co-founder Aymeric Vaudelin.
However, the team soon realized that the market wasn’t quite ready for standalone voice agents. To address this, they developed Donna, an AI voice agent tailored for anesthetists. While this might seem like an unusual niche, it was a strategic choice. Anesthesia secretaries in France handle a high volume of transactional calls—scheduling appointments, confirming availability, and addressing basic patient queries. These calls don’t require persuasion or complex problem-solving, making them an ideal use case for an AI agent.
Donna: A Proof of Concept for AI Voice Agents
Donna was Rounded’s first foray into the world of AI voice agents, and it proved to be a success. The agent was deployed in 15 private hospitals, handling hundreds of thousands of conversations. Early on, the team faced challenges with latency, with response times sometimes stretching to 4-6 seconds. But through iterative improvements, they reduced latency to under 700 milliseconds, making the experience nearly indistinguishable from a human conversation.
“With a web call, for example, we now get a latency of less than 700 milliseconds—more like 600 milliseconds. You add 200 milliseconds for phone connectivity, more or less,” Vaudelin explained.
Scaling Up: Rounded’s Orchestration Platform
Building on Donna’s success, Rounded has expanded its focus to create an orchestration platform that allows companies to build their own AI voice agents. The platform is designed to be flexible, enabling businesses to choose from a variety of off-the-shelf AI models for speech-to-text, large language models (LLMs), and text-to-speech. For example, a company could use Azure for transcription, GPT-4o mini as the LLM, and ElevenLabs for speech synthesis.
The platform also helps businesses define instruction trees and prompts tailored to their specific use cases. “Implementing an agent means finding the right prompting, the right parameters, and the right variables in the prompt,” Vaudelin said. “Our pitch is that we’re going to enable everyone to create great prompts or great agents, and our product will support them in this iteration process to create super-reliable agents.”
Why This Matters for Small Businesses and Entrepreneurs
For small business owners and entrepreneurs, Rounded’s platform represents a significant opportunity. AI voice agents can handle routine customer interactions, freeing up human employees to focus on more complex tasks. This not only improves efficiency but also enhances the customer experience by reducing wait times and providing instant support.
Moreover, the ability to customize AI voice agents means that businesses can tailor the technology to their unique needs. Whether you’re running a small e-commerce store, a local service business, or a healthcare practice, Rounded’s platform can help you create a voice agent that aligns with your brand and operational requirements.
Challenges and Limitations
While the potential of AI voice agents is immense, there are challenges to consider. One major limitation is the need for high-quality data to train the models effectively. Poor data quality can lead to inaccurate responses, which could frustrate customers. Additionally, while Rounded has made significant strides in reducing latency, achieving consistently low latency across different networks and devices remains a technical hurdle.
Another challenge is the ethical considerations surrounding AI voice agents. Businesses must ensure that their use of AI respects customer privacy and complies with data protection regulations. There’s also the risk of over-reliance on AI, which could lead to a loss of the personal touch that many customers value. For instance, in sensitive industries like healthcare, maintaining trust and empathy in customer interactions is critical, and AI must be carefully integrated to complement, not replace, human connection.
Key Takeaways
- AI Voice Agents are the Future: Rounded believes that voice agents will become the default way customers interact with companies, offering a more natural and efficient alternative to traditional methods.
- Customizable Solutions: Rounded’s orchestration platform allows businesses to build their own voice agents using a mix of off-the-shelf AI models, making it accessible for companies of all sizes.
- Improved Efficiency: By automating routine customer interactions, AI voice agents can help small businesses save time and resources, allowing them to focus on growth and innovation.
- Scalable Technology: With latency reduced to under 700 milliseconds, Rounded’s voice agents offer a near-human experience, making them a viable option for a wide range of industries.
As Rounded continues to refine its platform and expand its capabilities, the potential applications for AI voice agents are virtually limitless. For small business owners and entrepreneurs, this technology could be a game-changer, offering a competitive edge in an increasingly digital world. However, businesses must navigate the challenges of data quality, latency, and ethical considerations to fully harness the power of AI voice agents.
Leave a Reply