python machine learning guy

[login to view URL] Scoping and technical documentation [login to view URL] is the world’s first multi-industry regulatory compliance SAAS platform that helps businesses understand and create strategic decisions and plans around the regulatory landscape of their industries. The Compliance Engine is deep-learning system that not only acts as a database of laws, regulations, and statutes, but helps tie information together, helping create a map that businesses can navigate in order to gain and maintain compliance. Our primary focus areas are cannabis, cosmetics, and cryptocurrency. How the engine works: Regulatory framework library All laws, regulations, and statutes are inputted into a backend database. Text is parsed out and keywords relating to the laws are assigned. Every word within a law acts as a keyword, and broader descriptions and summaries of the law are created as well, acting as other links between regulations. Laws and regulations are added at the international level, state, territory, city, and municipality level in order to create as granular a data set as possible. Example: Cannabis packaging laws for California: [login to view URL] The information is entered in a raw form with the name of the law, section, and contents entered in unformatted text. From there we build out 2 additional fields 1. Summation field of that section where law is broken down into a 4-bullet point summary 2. A keyword field that pulls out the most important keywords from the law text to create symbolic links This creates the keyword database. For example, if someone runs a query for just “packaging” within the cannabis database, it will return all applicable laws from all jurisdictions. From there a user can refine their search to states and territories as need be. Once they click on that law, the result will then show the next steps of where to go. It will show in the process where the law sits. If a user clicks on a regulation of “vape pen” packaging, it will show them what they need to do before they can worry about vape pen packaging, and then where they should go from there. We also want this engine to then recommend all related laws based on filters that are linked to the keywords within that specific section of the law that the user is reading.   Backend: The database input needs to exist in a structure similar to this Airtable base: [login to view URL] It should be easy for someone to take the raw text of a law or regulation, assign attributes and categorize it, and then put the raw text in a field. It also needs to have the summary field and a keyword linking field so we can begin to group all relating laws by keyword. There needs to be a mechanism to link all keywords as needed and a filter system as well. Front-end: Easy to use user interface for quick searching, results that are visual, and represent a link between one another. Query results should be visually represented with links and checklists for the element that’s been queried. Example: A customer looks up “glycerin” The results should show all regulations regarding glycerin in the context of cannabis, from laws to ingredient percentages and allowed dosages. We’ll input all ingredients and dosages per statues and laws. Customer dashboard • Subscription type • Billable hours if applicable • Messages between customer and managed service account attorney • Recent queries