Vilnius-based nexos.ai has raised $8 million in funding, led by Index Ventures, with participation from Creandum, Dig Ventures, and several prominent angel investors, including Olivier Pomel (CEO of Datadog), Sebastian Siemiatkowski (CEO of Klarna), through Flat Capital, Ilkka Paananen (CEO of Supercell) and Avishai Abrahami (CEO of Wix). The company is developing an AI orchestration platform designed to help enterprises manage and optimize the growing complexity of AI models. As AI adoption expands, businesses face challenges in deploying and integrating AI efficiently while controlling costs and ensuring performance. nexos.ai aims to solve these issues by offering a platform that provides access to over 200 AI models from various providers, including OpenAI and Meta. The funding will be used to further refine the platform and address the increasing need for tools that streamline AI deployment across organizations.
nexos.ai was co-founded in late 2024 by Tomas Okmanas and Eimantas Sabaliauskas, the co-founders of several successful businesses including Nord Security (valued at $3 billion). The round attracted investment from prominent angels including Olivier Pomel (CEO of Datadog), Sebastian Siemiatkowski (CEO of Klarna), through Flat Capital, Ilkka Paananen (CEO of Supercell) and Avishai Abrahami (CEO of Wix.com).
As companies race to integrate AI capabilities, they face mounting complexity in managing multiple AI models, controlling costs, and ensuring reliable performance. The idea for nexos.ai emerged from Okmanas’ firsthand experience, where he struggled to integrate AI across several companies he was involved with – despite spending over $100k per month on large language models (LLM) usage in some cases. Conversations with other companies revealed the same pattern: businesses were eager to deploy AI, but lacked the infrastructure to create high-quality, secure, flexible and cost-effective applications at scale.
nexos.ai has solved this problem by building an AI operating system for enterprises, enabling them to harness and deploy hundreds of AI models through a single, seamless platform.
“Companies know that AI is an operational and competitive necessity, but they’re drowning in the challenges of managing multiple models, controlling costs and ensuring accurate and reliable performance,” said Okmanas, CEO and co-founder of nexos.ai. “At the same time, AI models are becoming increasingly autonomous and capable of handling complex tasks with minimal human intervention. We’ve built nexos.ai to be the enterprise-grade platform that makes working with AI as intuitive as working with human teams – providing the infrastructure and oversight to make sure these models perform at their best while remaining cost-effective and secure.”
Set to launch in Q1 2025, nexos.ai’s platform is already undergoing testing by major international companies for applications like automated customer support. The platform offers access to over 200 AI models from top providers such as OpenAI, Anthropic, Google, and Meta. It features key capabilities including smart model routing and load balancing to optimize both performance and costs, intelligent caching to reduce expenses for repeated queries, and comprehensive monitoring and analytics for AI operations. Additionally, it ensures enterprise-grade security and compliance controls, while providing automated performance optimization across various model providers.
“nexos.ai is a critical solution for any enterprise deploying AI at scale, where managing the inherently complex ecosystem of large language models, AI agents, and applications with varying levels of autonomy becomes essential. The demand for such solutions is expected to grow exponentially in the coming months and years,” said Hannah Seal, Partner at Index Ventures. “Tomas and his team have consistently demonstrated their ability to tackle complex challenges by creating products that deliver elegant, simple solutions — and this one might be their most consequential yet.”
Read the orginal article: https://arcticstartup.com/nexos-ai-raises-8m/