I have been asked by many startups to help them with their core machine learning IP
Unlike outsourced software development, machine learning is difficult to outsource or even to specify because it requires working deeply with the company data and product to invent trade secrets and even patents for the firm
There are very few machine learning professionals in the industry and the work does require experience and usually a PhD.
This raises the price and makes the contracts more complicated
What kind of contractual relationship would be ideal for both the startup to retain their IP and for the consultant to work with 2 or more clients at the same time, and over a period of time
That is, what is the best way for the firm to be sure their trade secrets wont be revealed, but also allows the consultant to work with his/her own developer stock of code and algorithms in order to solve related mathematical problems