Quote of the Day: “You rush a miracle man, you get a rotten miracle”
Miracle Max, The Princess Bride
When you start a new company, especially if you are trying to do something no one has ever done before, a common question you get asked is “who are you guys and where did you come from?”. (Another common question in our case is “Are you nuts? You’re doing what???” but we will cover that in another post.)
Back in the 1980s, I remember hearing an interview with the CEO of New York Seltzer; which was taking off like a rocket ship back at the time. The quote that stuck with me was when the CEO said, “We’re an overnight success story that was ten years in the making.”
Well, thankfully it hasn’t taken us ten years to get to where we are. Instead we are approaching our fifth year developing and perfecting the technology to provide fast, effective, and inexpensive anonymization solutions and now we’re bringing it to the general marketplace.
Our story started back in early 2017 when I left IBM. Under the heading of “Perfect Timing” I received a call from a past client who was the president of the board for the L.A. Trust for Children’s Health. I was eager for an interesting, and worthwhile challenge and did she ever have one.
As I was to quickly learn, the LA Trust oversees all of the healthcare related services provided to the 600,000+ students of the Los Angeles Unified School District. The LA Trust oversees more than a hundred medical, dental, mental and vision health facilities spread throughout the second largest school district in the country. What they needed right off the bat, was to get a grip on healthcare services being provided across the district. That sounded like a worthwhile project, but it was their next goal that got me hooked.
What they really wanted was to match the detailed healthcare related data from their independent healthcare service providers, with the detailed educational data from LAUSD, so they could start answering questions about which healthcare services had the greatest impact on the performance of their students. What impact on attendance, standardized test scores or graduation rates do different healthcare services have? Where, and in which areas, should budget be allocated? Billions of dollars are spent on healthcare for students in the US every year, but no one has ever been able to correlate the impact of healthcare access and treatment with academic performance.
Talk about a worthy cause. Miracle Max would be impressed!
As you can imagine, there were a lot of challenges. Chief among them was how would we ever afford the time and money it would require to build, and secure, a system which housed all of this highly sensitive, extremely risky and federally protected data? Those of you who are responsible for protecting PII know exactly what I mean.
We examined a lot of different options available to us at the time. Blockchain, differential privacy, various platform solutions were all examined but none of them were able to check all our boxes. So, we built our own.
The hurdles were many but with the combined talents and efforts of many dedicated people we were successful. For the first time ever, we were able to combine HIPAA protected medical records with FERPA protected academic records, at the individual level, and answer questions like “What happens to a student’s attendance once asthma treatment begins?”
Even better, our reporting database, where all this data is loaded, combined, processed, and aggregated does not have one byte of identifiable information in it. No PII was directly used to link this personal data and if anyone were to ever hack that system, all they would find is anonymous data. Every record, from every data source was anonymized independently, at the source, before it came into the reporting database where it is later combined and mined for these precious insights.
This is all made possible by the crown jewels of the L.A. Trust Data xChange; Privacy Environment or, as we call it, PII Vault Version 1.0.
A lot of effort was put into its design, construction and testing but as I look back, it is easy to see the most time was spent convincing all of the data sharing participants that this would work and that their data would be safe.
I can’t blame anyone now, and I sure did not blame them then, that we took so much time going through this. After all, our approach is so radically different from everything else out there that the inevitable question came up time and time again. “If this is so efficient, why has no one done it this way before?”
For those of you who have had to field this question then you know what it is like to come up with a new way to do something. If no one can provide a valid reason why it won’t work, then you know you have come up with a new and BETTER way to get something done.
And so it was, that in the end all of the security experts and attorneys brought in from multiple organizations, including LAUSD and Planned Parenthood Los Angeles gave, what we would later call Poly-Anonymization and Anonymized Data Matching, their blessing.
The L.A. Trust Data xChange is a success and a solution school districts around the country want to emulate and even license. The Privacy Environment, which easily saved us three quarters of the budget for the new analytics system, has been completely rewritten for the general market and is now the PII Vault from Anonomatic.
So that is how this started. A worthy cause, a dedicated team, and a worldwide problem that impacts every industry.
Intrigued if it will work for you? Feel free to try it out… for free.
Matt Fleck
Co-Founder & CEO – Anonomatic, Inc.