Machine Learning / AI
Since every implementation we do starts with accurate granular validated facts down to receipt-level detail (customer and product level), we end up producing an excellent source of ‘baseline’ data for machine learning logic.
We can help you push that data to your favorite machine learning tool and pull the results back in for reporting and analysis. Additionally, we can directly assist in defining and executing a project using our in-house consultants. We have a growing list of Machine Learning / AI partners and consultants should your requirements demand additional specialized tools and techniques.
Fact-Based
Our favorite Machine Learning project is one that uses a well-defined, nested “if-then” conditional rule to help with tasks like “Suggested Order” in supply chain or “Potential Voids” identification. We can generally do that all within our stack out-of-box.
Using Business Rules
Do you have business rules that take manual monitoring like every time a customer stops buying a ‘core product’ they should receive a visit/call type within 30 days.. or rules to identify out-of-stocks? We can help the computer find and surface those insights for you.
Market Basket, Forecast.. your choice
If you’re in retail and need to run the Apriori algorithm for market basket insights or in supply chain and want something more sophisticated than regressions, let’s define a project together.
Leverage “R” Python
Do you have staff who want to help?
We can set up a data stage for Machine Learning exercises using R and Python or your preferred code, database, and tools. Salient can write the code for you and perform the data exchange or just enable your internal team of analysts.
Custom Projects
While we’ve done some projects in consumer goods that can be repeated, generally every solution needs a little tailoring. We can help define business requirements and propose a solution.
Exchange Azure Data
Do you have a favorite app already like SPSS, Antuit, or the Azure Machine Learning Platform? We can help load baseline data to your tool and pull your results back in.