Lance Hampton

This gem can be plugged into a Ruby on Rails app (most useful before 4.0) to discover the routes at run time.

This is particularly useful when thrown into a large project and 'rake routes' takes a long time. Now a user can go to http:///route66 to view the routes immediately (since they are already loaded at run time.

At Numerex we had a recurring problem with fleet management that operators wanted to know what state their trucks were in, possibly for tax purposes.

Map services provide more data than is necessary in a reverse geo lookup (full address). They also take considerably more time when 10's of thousands of lookups may need to be done every minute.

This project solves the problem by breaking down states into triangles and finding which triangle contains the latitude and longitude provided. When used in library form locally it can process thousands of lookups a second.

The major work in this is the breakdown of states into their component triangles; the lookup being fairly straight forward. It has further been optimized to use bounding boxes to limit the search. Additional optimization has proven unnecessary at this time.

From their description:

"With the HeartAssistRemote™ Monitoring System, HeartAssist5® patients feel more secure about their heart health while enjoying life at home or traveling. The HeartAssist5® Conquest Controller continuously receives data from the HeartAssist5®. The data is transmitted in bursts to a controlled and secure data center. This data can be viewed remotely – helping to avoid unnecessary trips to the hospital. In cases of concern, HeartAssistRemote™ Monitoring provides a head start, allowing for proactive response and possibly better results for the patient. Effective deployment of this exclusive feature should result in better use of the caregiver’s time and the healthcare system’s resources."

This project was done with Ruby on Rails with messaging through the internet with failover to 4g. The monitoring was done on the device itself, providing clinicians and patients with alarms for early warnings or errors such as reduction in battery life.

As part of the Watson Healthcare initiative at IBM, consider regression analysis has been completed to track and predict the advance of disease based upon hundreds of potentials markers (lab results, physical characteristics, activities, etc.)

With the scoring system, this provides clinicians with the opportunity to reach out to those most at risk earlier. Ideally this results in prevention or diversion rather than handling greater disease later.

This project was completed largely in R and Java, with extensive backend analysis using a variety of analytic tools.

Redesigned and implemented a generic question/answer system that allowed all rules/questions/follow-ups to be configured by administrators.

This allows medical facilities to request specific requirements for their numerous types of users (doctors, drug reps, facilities, visitors, etc.)

The credentialing system enforces compliance in response and either facilitates the verification manually by credentialing agents or does it automatically (such as a background check).

This represented a move from custom code written for each type of facility. Now customization was moved to the onboarding team through customization, or even to the client themselves.

This project was completed using Ruby on Rails, PostgreSQL and extensive JavaScript. My specific role was largely as architect with hands-on development of the initial framework and the extensive messaging systems involved.

Start up company providing auto insurance quoted by the mile. We built a modular casualty insurance company and pivoted that into a software only company.

I developed the claims side of the effort. This included a model of the claim itself - types of damage, counting against which aspect of the insurance policy (think bodily injury vs. vehicle damage). From there the workflow enabled all levels to manage the claim as appropriate, enforcing authorization levels (management may be required to step in above a certain amount), tracking necessary steps (adjustment, legal) all the way through settlement.

Additional work was done to automate meet the state reporting requirements. This involved replicating the claims data stored in a Mongo database into a PostgreSQL database for reporting purposes; largely due to the difficulty in doing ad hoc queries across multiple conditionals in Mongo.

This project was done with Ruby on Rails, JavaScript, Mongodb, PosgreSQL, and Node.js.

IBM, having acquired multiple healthcare companies in a short amount of time has the problem of integrating multiple clients onto one system.

I designed and implemented a Single Sign On system using Golang to allow clients to have a single identity with their current credentials until they migrate over to IBM client IDs.