Contextual AI has recently launched and secured $20 million in a seed round led by Bain Capital Ventures with other key investors. The enterprise-focused AI startup, founded in Palo Alto by veterans of Facebook AI Research (FAIR) and Hugging Face, aims to develop artificial specialized intelligence (ASI). Douwe Kiela and Amanpreet Singh, co-founders of Contextual AI, stated in a blog post that they aim to focus on solving enterprise problems rather than wasting parameters, money, latency and compute on models that know Shakespeare and quantum physics.
The startup aims to address the six problems that society faces through the use of current Large Language Models (LLMs). These problems include the lack of domain-specific knowledge in existing LLMs, the expensiveness of fine-tuning and training LLMs, the difficulty in scaling LLMs to lower-resource languages, the high churn rate of open-source projects, the potential risk of using commercially-available LLMs and the lack of detection of unreliable information and misinformation.
Contextual AI encourages the use of open-source software and pledges to ensure more players are involved, rather than only incumbent big-tech corporations handling the technology. The company has 12 engineers so far, with the two co-founders having expertise in AI and LLMs. They believe that the firm’s technological approach supports open sourcing, and the community will leverage this by providing more open source software and playing an active role in the community.
BCV Principal Rak Garg and Partner Aaref Hilaly have expressed their support for Contextual AI’s goals and ambitions and believe there is significant potential in the market for enterprise-grade AI. They believe that AI will become ingrained in everything enterprises do in ways that are secure, deeply integrated and, most importantly, contextual to employee workflows. They are looking forward to working with Douwe and Aman to bring the power of language models into the workplace.
