Contextual AI

Contextual AI is a Silicon Valley-based startup founded in 2023 by Douwe Kiela and Amanpreet Singh, former researchers at Meta and Hugging Face respectively. The company specializes in developing advanced retrieval-augmented generation (RAG) technology to enhance the performance and accuracy of large language models (LLMs) for enterprise applications. Contextual AI's flagship product is RAG 2.0, an enterprise AI platform that enables companies to build state-of-the-art RAG applications quickly and efficiently. The technology allows LLMs to access and incorporate external data sources, resulting in more accurate and context-aware responses.

Contextual AI's approach focuses on two key principles: Systems over models and specialization over artificial general intelligence (AGI). By optimizing the entire system end-to-end, including the retriever and language model architectures, RAG 2.0 claims to achieve up to 10x better parameter accuracy and performance compared to competing offerings. This optimization allows for a more efficient use of computational resources, enabling smaller infrastructure to run models that typically require much larger compute capacity.

The company's platform is designed to be LLM-agnostic, working with various open-source language models such as Mistral and Llama. Contextual AI developed its retrievers using NVIDIA's Megatron LM on a mix of NVIDIA H100 and A100 Tensor Core GPUs hosted in Google Cloud. The platform can be deployed in the cloud, on-premises, or fully disconnected environments, making it suitable for various industries, including FinTech, manufacturing, medical devices, and robotics. The company’s business model is on a Saas basis. 

Key customers and partnerships 

HSBC has partnered with Contextual AI for AI-assisted knowledge management. In August 2023, Contextual AI partnered with Google Cloud to bring GenAI to the enterprise.