Customer case study - LinkOrb

I’ve been silent for some time as I was busy working with this client, but as a reward for your patience I will now take you on a journey. Join me to see what the problems this organization faced were as well as the solutions applied.

Overview of the business Link to heading

LinkOrb is a Healthcare IT company, their vision is to empower individuals with autonomous ownership of their data, fostering a culture of trust through easy and transparent data exchange. They are dedicated to enhancing the well-being of society by facilitating efficient healthcare solutions that promote the good health of individuals Mission

  • Empower expectant parents to experience a well-informed, well-prepared, and worry-free pregnancy journey through the delivery of trustworthy, relevant, and secure information and communication.
  • Enable healthcare providers to deliver high-quality, efficient, and joyful care.
  • Facilitate advertisers' reach to their target audience, both regionally and nationally, through the largest and most active pregnancy community.

Products Link to heading

Some of the most popular solutions provided by LinkOrb include, but are not limited to:

  • Two Electronic Health Record (EHR) software systems (one acquired recently). Used by healthcare professionals, specialized in pregnancy care and child birth. More than half of Dutch babies are brought into this world with the help of these systems.
  • A patient portal guiding over 2.5 million new young parents through their prenatal journeys.
  • A DICOM solution used to exchange thousands of ultrasound images and biometry reports daily. The most complete medical data exchange platform used by leading hospitals and surrounding healthcare professionals.

Challenges Link to heading

As can be seen from the overview the LinkOrb team isn’t wanting in ambition, nor the ability to execute on it from a software engineering perspective. Throughout their years of growth and expansion, however, they did begin to face infrastructure hurdles:

  • Over 100,000 monthly alerts from monitoring systems, predominantly of a critical and urgent nature.
  • Almost daily occurrence of critical issues causing major downtime for the product offerings and backlogs of other tasks and general infrastructure improvements
  • Systems that required near constant oversight in order to remain operational at an expected level - Backups, SSL, system updates etc. In many ways these are pitfalls that are common for modern day organizations and can become a serious roadblock to reaching their full potential. While many of the building blocks were there, in order to construct the necessary solutions the team needed the help of a trusted partner to get them to the desired level of autonomous systems.

The solutions Link to heading

As touched on in my other writing, monitoring tools are a tremendous source of insight nowadays and luckily they had implemented a modern monitoring solution, so the work was focused on reviewing that at first. Getting hundreds of thousands of alerts can cause a “missing the forest for the trees” type of scenario, therefore it was critical take a step back and adopt “big-picture” approach.

After conducting manual meta-analyses of the output from their monitoring systems, so that the main pain sources were identified and proper priorities established, the next step is to actually address those root issues where the highest source of actual, not system reported, end-user pain is first on the list. I can’t disclose too many details of what was actually done in terms of root-cause resolution here are some highlights:

  • Re-assessed and re-configured load balancers for optimal performance and minimal downtime.
  • Optimized server resource utilization ensuring proper availability and costs.
  • Multiple enhancements to the backup platforms:
  • Reconfigured backup system to handle multiple servers simultaneously, reducing backup queue and resolving missed backup alerts. 
  • Introduced a new backup solution capable of handling large file volumes for servers with extensive data where the old system couldn’t actually handle them in a consistent manner.
  • Bring systems in line with latest software releases, so that newer features for stability were available.
  • Updated automation freeing up resources for other priorities.
  • Constant adjustment of the monitoring systems to reflect reality and real prioritization of problems. There are aspects which require ongoing work, but even now some very good results are evident from this approach:

Significant Reduction in Alerts: Link to heading

  • Decreased alerts by 90%, shifting focus to lower-priority issues.

Improved System Stability: Link to heading

  • Downtime due to website issues minimized. 
  • Backup processes streamlined and effectively managed. 
  • SSL issues resolved through systematic updates and maintenance.

The monitoring meta-analysis, used as a benchmark for performance, helps shine a light on what has been addressed and points the direction on what requires work next. That also enables clear communication with the LinkOrb team and helps set expectations. As it is so important, I also automated it and created a tool (source code is available here, per my love for OSS), which returns weekly and monthly results. There is further work to be done on extending it with APIs which can be plugged into other tools for even better data visualization.

Conclusion Link to heading

Using a more data-centric, scientific mind and skillsets, organizations can adopt the big-picture approach, described above, so that large improvements can be brought about. Usually when systems are initially designed the availability of data isn’t taken into consideration, as there are constraints in terms of time and other resources. This is not optimal as whenever a system grows to meet business needs it becomes even more difficult to make informed decisions based on individual understanding.

With LinkOrb a different facet of the perils an organization might face along their data journey was made apparent. Their platforms had the prerequisite level of data generation, but it was not being utilized to it’s complete potential. This wasn’t because they hadn’t considered it a critical business need, or a cultural problem. A loss of focus, centering attention on single problems and falling into toil type of habitual issue resolution, obfuscating root causes and making them difficult to address are as dangerous as not having the data in the first place.

With this methodology, established resolutions, and validated results, LinkOrb now possesses all the necessary resources to support healthcare providers and enhance the pregnancy experience for expectant parents, ensuring it is informed, prepared, and stress-free.

I wish them many years of trouble free operations as their journey is a noble one!