I help creative entrepreneurs develop Machine Learning Products.

Innovate with Assurance. Mitigate the risks of your initial investment in machine learning by working with an expert. I've helped several companies translate novel concepts into high performance machine learning systems quickly, reliably, and affordably.

Working with me

Discovery Meeting. We have a chat about your business problem, how it might be translated into an ML problem, and what data sources we can use to solve it.
Defining Success. We choose a metric for success that will be used to assess the performance of the model. We also choose the minimum level of performance that would be acceptable in your application.
Data Audit. I inspect the data available and assess whether or not current data sources are sufficient for our purpose. If not, I'll make recommendations on ways to reframe the problem.
Baseline Model. I create a simple model and test its performance to validate the feasibility of the task given the data.
Model Development. I develop a model that performs as well as possible given the time & resource constraints. At the conclusion of this process, the final candidate model is tested for acceptability.
Deployment & Maintenance. If the model passes the final acceptance test, we build the software needed to enable the model to be used by end-users and by other programs. Once launched, we proactively monitor how the model performs in the real world and keep it working as expected.

About Fisher Metrical

Bringing your ML Product from Concept to Market.

Many valuable machine learning use-cases present challenges for which a google search will not return easy or complete answers. As an entrepreneur, standing in such unexplored territory offers genuine promise but also plenty of enticing illusions and unforeseen risks. I can offer the open-mindedness, careful listening, and technical insight needed to help you make the translation from a concept to a feasible technical strategy. From there, I can help you develop reliable, statistically sound, and efficient machine learning software. Let’s jump on a quick & complementary video call to go over your use-case.

“When it comes to applied machine learning, the hardest part is knowing what you want to cook and how you plan to check it before you serve it to your customers.”

- Cassie Kozyrkov