The Trade-Off Between Precision and Recall in Predictive Networking

Send us a text

Cisco got its start in 1984 connecting computers at Stanford University to form the first local area network. Other than maybe Microsoft or Apple, it’s easy to argue Cisco has had more influence on the growth of the internet, and by extension, the modern world, than any other company. 15 years after Cisco started today’s guest was hired to begin what would become a legendary career. 

Nearly 25 years later JP Vasseur has changed the world again and again. In the process, he has been recognized as the #1 inventor at Cisco with 600 patents to his name. He has authored or co-authored 35 standards, published three books on internet technologies, and has been recognized as a Cisco Fellow, a prestigious title awarded to the top few most-distinguished technical leaders at the company. Today we learn from a living legend about the past, present, and future of technology.

Listen and learn...

1. How AI at Cisco has evolved in the past 12 years
2. Disruptive vs. incremental innovation
3. How predictive networks learn
4. The design principle JP used when designing the first predictive network
5. The challenges of predicting outages using unsupervised vs. supervised machine learning
6. JP's process for innovating like a startup within Cisco
7. Innovation in networking we can expect in the next decade
8. JP's best memory from the early days of Cisco

References in this episode:

• JP's blog
• Kevin Roose from the New York Times had a disturbing conversation with Microsoft's Bing
• Chambers Talks, the great podcast from former Cisco CEO John Chambers
• Yann LeCun on how babies learn
• Phil McKinney, former HP CTO, on AI and the Future of Work