Loading...

We hear the word AI get thrown around all of the time. AI this, algorithm that. "Look at this cool new way AI is being used for another completely useless thing". Like many, I wondered what some actual real-life use cases are that make sense for involving machine learning. This is not an exhaustive list, but here are some ways that machine learning can be helpful.

1. Text file data extraction and pattern recognition for document types

Have a set of documents that you need to wade through? Judging by file names and manually opening files can get very tedious, very quickly. It can also be error prone. The larger the data set, the more this will become clear. One way to handle this is by recognizing the document type using machine learning and smart labeling. Once a portion of a given set of documents is labeled, it can be interpreted by a machine on another set. A good rule of thumb for where this begins to be most useful is when there is a dataset of at least 10,000 documents.

2. Image recognition for matching copyrighted photos

Have a lot of copyrighted images? Say you're a publication with photos belonging to you, or a photographer doing a lot of shoots. Machine learning tools can be used to match up your copyrighted photos with those on other websites. While the web is a very large place and searching its entirety can be more effort than it's worth, focusing a subset of sites where it happens frequently can be a great place to start.


3. Numeric statistical prediction for sales teams

Manage a large sales team and want to better understand what factors truly drive sales? Tools like Hubspot and Salesforce (and other CRMs) have lots of statistical data to work with. Proposal sent? Meeting set up? These are all categorical data that can be used to drive a strong statistical forecast of what the most likely result is. This can result in significantly improved forecasting for large sales teams.


Want to learn more about practical machine learning for businesses? Feel free to contact us below.


Want to work with us?

We are always open to interesting projects.

Work With Us