What happened
Jeff Bezos has backed Profluent, an AI startup that uses advanced AI models to design programmable proteins. Profluent recently raised $106 million to expand the application of its AI-driven protein design technology in drug development and agriculture. The company leverages a vast protein dataset and proprietary AI models to create novel proteins with potential real-world uses. Its Emeryville, California headquarters integrates AI engineering with wet-lab testing to validate new protein designs.
Why it matters
- Profluentâs AI platform could transform medicine and agriculture by enabling the design of proteins tailored for specific therapeutic and industrial purposes.
- The startupâs technology represents a significant advance over traditional protein discovery by using large-scale AI models trained on the worldâs largest protein sequence catalog.
- This funding round supports scaling the use of AI in programmable biology, potentially accelerating drug development and biomanufacturing innovations.
Key facts
- 1. Profluent raised $106 million to scale its AI models for protein design.
- 2. The company was founded in 2022 and is based in Emeryville, California.
- 3. Its AI models, including ProGen3, are trained on over 3.4 billion protein sequences and a protein catalog of 115 billion unique sequences.
- 4. Profluentâs dataset includes about 20 trillion amino acid tokens, surpassing the data volume used for training models like AlphaFold 3.
- 5. The AI platform enables single-shot design of antibodies and ultra-compact gene editors targeting millions of patients and large market opportunities.
Key Voices
- Ali Madani, Profluent CEO, emphasized the necessity of combining AI with real-world laboratory testing for effective protein design.
- Profluent highlights its unique approach of using bespoke AI models trained on the largest protein sequence database globally.
- Not stated in the article.
