Machine Learning & AI
We don't try and sell you a specific AI solution. Here's how we approach machine learning & AI.
Use Case
We honestly wouldn't touch AI or any Data project unless we had a use case. There has to be a reason to do something
Full Scope & Technical Requirements
We define the inputs, the data cleaning required, and the outputs.
Train & Test
We train our ML Models on 80% of data and then test it on the other 20%.
One of the crucial components of an ML project is knowing WHAT to optimize on. We create a baseline based on the exact metric that is most relevant to the use case
Optimize
There's a fine balance in ML Model Optimization. You don't want the model to memorize behavior.
You do want the model to learn and make predictions for future data.
Deploy to Production
Once we're confident in the predictions and the accuracy obtained, we deploy into production
What do our customers say about machine learning?
"I just wanna acknowledge how f**ng legendary you've been navigating this situation. Your quick response time, willingness to work with us, your creativity... seriously A+ man"
^ this was from a client who we deployed an ML model to, that created daily predictions of the behavior of certain fields in a dataset
Pricing
Pricing can vary widely. In general, small ML projects with existing relatively clean data start at around $20,000.
Large projects, and projects that require a lot of data cleaning, could be year-long initiatives with full time resources.