What can you learn from how they’ve been doing it?
Build and harness “pockets of talent”.
For every one of their machine learning incubations, Flagship has been able to find people that have deep perspective on how to actually how to actually do the work of applying machine learning to biology. The interesting thing is that this talent cultivation is not merely incidental, it's intentional. Flagship has made it a point to harvest the Boston and Kendall Square scheme for its talent. They have also followed a clear strategy for garnering ML talent.
They started a fellowship for students in machine learning.
They have thoughtfully developed templates for hiring the software engineering and machine learning talent required to productionize ML. (Sample JD)
They hired and empowered superstar translational talent in computational health like Andrew Beam.
They have intentionally fostered connections amongst machine learning-oriented portfolio companies.
Future company builders should do the same: identify, develop, and harness unique pools of talent.
And it's continued to draw from that same well as machine learning has grown in prominence in that community. And the very, very smart people that may have been doing biology work before are also starting to think about the application of machine learning in that context. Let's take a look at each of the companies and how they're actually applying machine learning to their platform strategy. In order.
- they have an incredible collection of talent
- drawing heavily on the boston and kendall square scene
- umut eser, milind kamkolkar
- andrew beam, jon bloom, jess johnson
- avak, molly, geoff maltzahn, david berry, armen mkrytchan, karim lakhani
Patience truly is a virtue, especially for platform.
Your investors will never have patience.
Patience is NOT a virtue. Patience is the intentional cultivation of a suboptimal short-term outcome for a disproportionate long-term outcome.
Thinking in arcs is a is a powerful skill that requires tremendous patience. And that is the part that people underestimate. Right? Is the patience required to see an arc through through its troughs and its peaks and the importance of timing. None of these lessons are new. But the way we teach ourselves is through applying them to the examples that we see in our lives. And that's the way I'm basically doing that here.
Embrace a vision.
harness or create alternative sources of data.
you dont just build the algorithm. you build the dataset.