The UK-based voice AI startup SLNG has introduced a platform aimed at reducing costs within the voice agent industry, which has reportedly seen inflated pricing due to compute-heavy business models. This initiative emerges as a response to the company’s observations that the current market landscape encourages excessive use of computational resources, leading to rising operational costs for enterprises relying on voice AI solutions. Operating across 11 regions, SLNG capitalizes on region-specific compliance requirements, positioning itself as a comprehensive solution for companies globally.
SLNG’s unique execution layer for voice agents is pivotal to its strategy, accommodating existing frameworks like LiveKit and Pipecat by managing AI model selection, routing, and compliance. This setup contrasts previous industry practices where companies often had to customize their infrastructure. The execution layer SLNG offers address these challenges, streamlining the deployment processes and cutting both latency and costs by around 50%.
How does SLNG’s execution layer work?
At its core, SLNG’s execution layer efficiently coordinates processes like speech-to-text and text-to-speech, ensuring optimal model selection only when high-caliber computations are necessary. Luke Miller, the CEO of SLNG, points out that this tailored approach not only reduces expenses but also enhances performance outcomes like appointment confirmations and customer resolutions. By focusing on these optimizations, SLNG offers a more consistent and reliable platform for handling high volumes of calls.
Why has SLNG focused on global compliance?
SLNG’s focus on global compliance fills a significant void in markets such as Southeast Asia, Latin America, and India, where traditional models of computational abundance seen in the US are not applicable. This adaptation to global contexts allows SLNG to meet stringent data sovereignty laws in sectors such as finance and healthcare. As co-founder Ismael Ordaz highlights, this discipline of making do without excessive GPU availability has become central to SLNG’s value proposition.
Evidence of SLNG’s impact is demonstrated by clients like Ixigo, an Indian travel operator. Previously managing numerous vendors, Ixigo now consolidates its voice AI operations through SLNG’s platform, simplifying its infrastructure needs and allowing for a more focused engineering approach on end-user satisfaction and outcome improvement.
The company’s ongoing infrastructure expansion supports its aim to democratize voice deployment similar to how AWS and Vercel have transformed compute and web deployments respectively. This objective not only broadens SLNG’s market reach but also establishes a regulatory advantage in Europe, a region with extensive compliance mandates.
Looking towards the future, SLNG foresees its environment as the normative platform for developing voice agents, catering to both human developers and AI entities. “Voice is how humans interact, and the execution layer needs to be ready for agents to create that interface on-demand,” emphasizes Luke Miller, outlining SLNG’s vision for the AI agent-driven future.
