Dr. Sara Hooker
Dr. Sara Hooker, Head of Cohere For AI, a research lab that seeks to solve complex machine learning problems.
Did you know it was a woman who brought Google’s first AI research office to Africa?
Meet Dr. Sara Hooker, VP of research at Cohere – the enterprise AI platform last valued at $2.2 billion!
After completing a double major in Economics and International Relations, Hooker took up her first role as a Senior Economics Analyst at Compass Lexecon. It was years later that she obtained her Ph.D. in Computer Science from the Mila-Quebec AI Institute.
In 2014, she founded Delta Analytics, a non-profit where data analysts would volunteer to help other non-profits use data to further their impact. (With no prior education in computer science, Delta Analytics gave Hooker a gateway into data and tech!)
In 2015, she joined Udemy as a data analyst and developed the startup’s first spam detection algorithm. She also took part-time evening classes through the non-profit fast.ai and completed a 12-week course on AI.
In 2017, she was accepted into Google’s AI Residency Program, where she served as a Research Scientist till 2022. Here, Hooker became a founding member of Google’s AI research office in Ghana.
(She achieved all this before pursuing her Ph.D – as I say, you don’t need to be a techie to have a seat at the tech table!)
In 2022, she was appointed to lead Cohere For AI, the company’s research lab, which focuses on making research available to the open-source community.
Hooker is the co-founder of Trustworthy ML Initiative, a forum related to Trustworthy ML. She is on the board of several councils including:
→ World Economic Forum’s Council on the Future of AI
→ Kaggle's ML Advisory Research Board
→ MLC’s research group
“My unusual background also brings certain advantages…[it] often leads me to connect concepts in unexpected and novel ways.”
𝐅𝐮𝐧 𝐅𝐚𝐜𝐭: Sara Hooker hosts a podcast called 'underrated ml’ alongside her brother Sean, where the two invite different guests and discuss ML research papers. At the end, all viewers are invited to vote on which paper they found the most underrated.
Last Updated: May 2024
