Doubleword

Doubleword, formerly known as TitanML, is a London-based company founded in 2021 that develops a self-hosted inference platform for enterprises. The company was founded by Meryem Arik (CEO), Dr. James Dborin (CSO), and Dr. Fergus Finn (CTO). The founding idea originated from the postdoctoral research of Dr. Dborin and Dr. Finn at University College London (UCL), where they discovered that techniques from their work in quantum machine learning could significantly improve AI inference performance. Arik, an Oxford alumna with a background in theoretical physics, joined them to co-found the company.

Doubleword addresses the challenges businesses face when deploying large language models (LLMs), such as high operational costs, infrastructure complexity, and data privacy concerns. The company provides a platform that simplifies the deployment and management of AI models, including open-source, proprietary, or fine-tuned versions, on any hardware, whether in the cloud or on-premise. This allows clients to maintain control over their AI applications and data. Their flagship product, the Titan Takeoff Inference Server, is designed to make LLM deployment faster and more cost-effective, demonstrating up to 90% reductions in compute costs and significant latency improvements for clients. The platform supports a wide range of models, including generative language models, embedding models, and image-to-text models.

The company's business model is B2B SaaS, targeting enterprises, particularly in regulated industries like finance, that need to self-host AI for security and compliance. Doubleword has formed strategic partnerships with companies like Intel, AWS, Snowflake, and Dataiku. In October 2023, while still operating as TitanML, the company raised $2.8 million in a pre-seed funding round led by Octopus Ventures. In May 2025, the company rebranded to Doubleword and announced a $12 million Series A funding round led by Dawn Capital, with participation from K5 Global and notable angel investors.

Keywords: self-hosted AI, large language models, AI inference platform, model deployment, enterprise AI, MLOps, AI infrastructure, model optimization, model compression, secure AI, private AI, natural language processing, TitanML, Titan Takeoff, deep learning, AI for finance, data privacy, cloud deployment, on-premise deployment, AI scalability