Chatbots, messaging apps, and other conversational user interfaces are gradually becoming the norm. Though it is our aspiration, we don’t typically utilize spoken dialog in our software (it’s still too costly except in special cases).
However ‘conversational’ also means to us an interface that is simple, self-evident and forgiving of human limitations. We use natural language processing (NLP) where possible to simplify the user experience.
In a world flooded with data, we need help to make sense of it, find insights, and visualize relationships in a way that promotes understanding, decision-making, and real-team interactions. We utilize an array of techniques to go beyond record-keeping to software that classifies, predicts and recommends.
Text mining, word vectors, association rules, principal component analysis, correspondence analysis, k nearest neighbors, decision trees, neural networks, and regression are among the algorithms we use. Ultimately it means software that “knows the story inside the data”.
Automation has been around for a long time, but autonomous software is fairly new. It’s software that you direct to “do it for me”. Smart Devices (aka “the Internet of Things” IoT) allow software to sense the world, reducing the need for human attention. We make software that is self-regulating, sensing, interpreting, and able to take action in prescribed ways based on understood patterns.
Mental Health Apps
We develop software that helps clinicians (therapists) work with their clients: it monitors, evaluates and visualizes a client’s psychological state. Borrowing methods from Cognitive Behavioral Therapy (CBT), Appraisal Theory and the Self-Efficacy literature, we design the software to “connect the dots”, and to construct a composite portrait of the individual. By gathering a person’s emotional and cognitive state, the software facilitates treatment by a trained psychologist. (All data is secured by industry-standard encryption.)
Medical Facility Management
Operations at all medical facilities move at a rapid pace. Whether the ER or an elective surgery clinic, medical procedures are taking place in a continuous cycle with doctors, patients, nurses and equipment continuously in motion. EMR systems address primarily the billing needs of these complex systems. We work with sensor hardware and cloud storage systems to manage vast amounts of real-time data to create alerts, visualizations and reports in order to optimize the flow of patients through these facilities.
We use a variety of software building blocks and web services (cloud API’s). Our typical configuration:
Front End Development: Shiny to create responsive web applications (works great on mobile, tablet or desktop)
Predictive Analytics: R and RStudio to develop sophisticated data models, analytics and visualizations
Secure, Flexible Database: MongoDB is our db of choice to manage transactions and store data repositories.
While this is not suited for every use case, we find that by sticking to this simple configuration in most cases, we can keep costs down and deliver powerful, useful Software as a Service to our customers at a great value and rapid development.