Creators of Transformer Architecture Raise Nearly $100M for New AI Startup Keen Technologies

Keen Technologies, a new AI startup founded by key architects of the revolutionary Transformer model, has emerged from stealth with a significant infusion of capital. The company, established by former Google researchers Niki Parmar and Ashish Vaswani, has successfully raised nearly $100 million in a funding round led by prominent tech investors Nat Friedman (former CEO of GitHub) and Daniel Gross.

Parmar and Vaswani are two of the eight co-authors of the seminal 2017 research paper, “Attention Is All You Need,” which introduced the Transformer architecture. This technology became the foundational building block for virtually all modern large language models (LLMs), including OpenAI’s GPT series and Google’s Gemini. Their departure from Google and the formation of Keen Technologies marks another significant moment in the exodus of top AI talent from established tech giants to build more agile and focused ventures.

The San Francisco-based company describes itself as an “AI research and product lab” with a mission to develop new foundational models and user-centric products. While details about their specific architectural innovations remain under wraps, the founders’ pedigree suggests an ambition to move beyond the limitations of the very technology they helped create. The funding round also saw participation from investors like TCV, Lightspeed Venture Partners, and others, signaling strong confidence in the team’s vision.

This development places Keen Technologies in an intensely competitive landscape, challenging established players like OpenAI, Google, Anthropic, and Mistral. The move mirrors the formation of other successful AI startups by Google alumni, such as Cohere and Character.ai. As the AI industry continues to consolidate around a few dominant architectures, the launch of a well-funded lab led by the original creators of the Transformer signals a potential paradigm shift and a new chapter in the race to build more advanced artificial intelligence.

Leave a Comment

Your email address will not be published. Required fields are marked *

en_USEnglish
Scroll to Top