Products

What It Does:

Matchmaker produces a master index and linkage table to connect records between databases.

Context:

Multi-service organizations often rely on several software systems and databases to deliver their work. This is because one tool usually can't do it all. For example, a system designed for scheduling meals on wheels may not be optimized to support another specialized service like employment counselling. In other cases, a funder may require that specific services use a central database.

The Problem:

In these settings, a participant that uses more than one service may wind up in multiple databases. As a result, it is difficult to count how many unique individuals are being served by the organization. And because participants are not usually tracked across systems with a common identifier, it is impossible to understand the pathways with which they are moving through an organization.

How It Works:

Matchmaker is built using DataMade's excellent dedupe Python package which employs fuzzy-matching techniques and cleverly leverages machine learning to cluster and weight matching criteria. The end result is a table with a single unique ID for each individual and a list of IDs for the corresponding source databases. If desired, this table can be used to create a deduplicated, consolidated database of unique individuals served by the organization.

Tech Specs:

Built in Python and compiled with a user interface for Windows, this program runs locally so that sensitive data never leaves your secure environment. It can also be adapted for other operating systems or automation on a server.

What It Does:

Retriever is an automated tool that fetches reports from web-based client management systems.

Context:

Client management systems excel at helping organizations to record and track information about service users. They typically come with a limited set of "canned" reports, but when these reports don't cover all your needs, the next best option is to export the data and build your reports in another tool.

The Problem:

Some client management systems can be greedy: they readily consume your information, but make it difficult to get it out. For example, exporting data can be a manual process that requires users to select a report from a list, set date ranges and other filters, and select a download format and destination. It's not a lot of clicks, but it adds up if you want to update your data daily or if you need to export multiple reports at a time.

How It Works:

Retriever mimics your mouse and keystroke interactions to automatically complete the steps required to export data from your web-based client management system. It can be set up to run on-demand or run on a scheduled basis (e.g. nightly) on a server. Choose as many reports as you like and define the date ranges and filters to configure for each export. The end result is a folder of reports saved in an Excel-friendly format that can be downloaded with a single click of a button.

Tech Specs:

Built in R Shiny with Selenium Server, Retriever has been programmed for use with VitalHub's Pirouette, but can be adpated for other web-based client management systems. It provides a simple user interface that can be triggered manually or set to run automatically on a server.