data mining

Spotlight – Tragedy Produces Crime

Liability cases can be tricky. Worked on a case once that turned from a AOE/COE into a subrogation once we uncovered the cause of death was a faulty support system installed by a 3rd party. Sadly, the death occurred just before Christmas. Another case involved a man who sued a major car maker because their car caught fire due to the man’s DUI influenced wreck. Still another saw a man lose his wife, of fourteen years, and his four children when the SUV exploded.

Yes, the cases are often sad. And sometimes, they include a twist. Like the time a malpractice lawsuit turned into a mafia operation. No joke.

Picture this: man goes into hospital with chest pains. County hospital misdiagnosed him (I’m not sure how they didn’t suspect a heart attack), and he dies. Family sues the country for a few million. I am tasked with checking in on businesses and assets on the deceased and his immediate family.

Here’s where it got weird. Ownership of all the businesses, all of which were involved in medical supplies, were being transferred to family friends. This included all the businesses which should have been owned by the subject’s spouse and children. Nope. Every business was transferred, and even the property deeds were being toyed with.

This case became so confusing I had to generate a data model to show all the relationships and connections between family and friends (including phone numbers, addresses, etc.). I also ended up doing bank searches and spot checks. I found some of businesses even changed their name or registered out of state.

The funny thing? They weren’t asking for more money. No, everything was being done to keep prying eyes away from their books. Allegedly, the entire thing was some sort of mafia racket. I didn’t stay on this case long enough to see it through. I handed over my findings to the county and we all went our merry ways. One of the craziest plot twists ever. Not the most insane though. I’ll leave that story of the vegetable turned real estate mogul for another day.

Case Study – Data’s Little Bro

Aside from fighting off guard dogs, hiking seven miles a day and navigating backyard death traps, entering accurate data into a computerized device played a key role in meter reading. I should know since I did it for about four and a half years for one of the largest utility companies in the world. It served as my first taste of data entry, but it wouldn’t be my last.

To this point, in my case study series, I’ve discussed a sort of data family: database administration, data mining, data mapping, and data analytics. When people think data, they think data science. What people forget is the importance of good, clean, accurate and efficient data entry. Entering accurate data into a system is an essential task. And I’ve seen far too many get it wrong, whether at the corporate or government level.

When I started working at an investigative firm as a data entry associate, I was met with a curve ball: it wasn’t just data entry. It was comprehensive data entry. I had to comprehend the request I was being asked to enter into the online database. In other words, typing speed was pointless. Accuracy, speed, and thoroughness was everything.

So, I did what I knew best: I asked a ton of questions. I researched the insurance industry. I learned the ins and outs of claims. I broke down “referrals” according to case type, and ultimately developed my own system for maintaining accuracy and speed.

Within a week, I was already hitting double digits a day when the goal had been eight. It wasn’t long before I set records for most done in an hour, a day, a week, a month, a year, and a lifetime for speed and accuracy. I wrote the training manual on the department I wound up supervising for a spell.

I was never the fastest typist. Nor was I the most experienced in insurance. I simply worked my tail off in order to completely transform how an entire department operated.

Never be afraid to take a task others find unimportant and tedious to new heights. Work hard. Stay strong. Set records. Be faithful in the little things. And lay down solid foundations.

Case Study – Intelligence / Investigative Analytics

Over the last seven years, and for more than 12,000+ investigative hours, I’ve worked as a intelligence/investigative analyst. I use both intelligence and investigative because the position goes by either.

Technically, I’m an All Source Analyst because I use both open (OSINT) and closed source data points. An intelligence analyst researches, gathers, and evaluates data from a variety of sources. They specialize in data mining. An investigative analyst works on, you guessed it, investigations.

Typically, there are three divisions: military, law enforcement, and insurance. Often, we use similar databases but with different levels of security clearance. Law enforcement is more concerned with digital forensics and cyber investigations. Military locate and track terrorists. Insurance investigates claims. All three have similar skill sets: we are cyber sleuths. It should be noted that there are also Cyber Security Threat Analysts that are also similar (they search systems and networks).

The insurance side of things deals a lot with insurance companies and law firms, and often works alongside law enforcement. Some of the cases I’ve worked on include money laundering, rape, assault, kidnapping, and shootings. Several of them have gone international (Mexico, Spain, Canada, Guatemala, etc) , and have included all forms of insurance: liability, work comp, property loss, FMLA, and life.

Before I continue, it should be noted that a background investigator is not the same as an analyst. Yes, I may do a simple social media sweep or a pre-employment search, but that’s the extent of what a background investigator does. They collect a bunch of data, but make little to no attempt at evaluating it. They don’t go beyond the confines of a search engine. Additionally, a copy service retrieves documents such as court records, birth certificates, and camera footage from an intersection, but that is only the “other” duties of my job function, and not what I do the majority of the time.

So, what do I do? I piece puzzles together in order to paint a clearer picture. I work on skip traces, SSN traces, heir searches, asset and business searches, employment checks, social media archives (which may include metadata) and criminal/civil checks, and a host of other case types. I launch bank account searches, comb through DMV records, and run vehicle sighting reports. I might triangulate the location of a cell phone or create a family tree in Ancestry.

Yep, I’ve read thousands of police reports, traffic incident reports, birth certificates, property deeds, property transfer detail reports, vehicle titles, death certificates, autopsies, bankruptcy documents, articles of organization and incorporation, statements of info, and marriage and divorce records. I conduct geofences, match data points, and watch body-cam and surveillance footage. I then conclude my findings by compiling a legal document for court purposes.

My toolkit is vast. It includes a host of software and online databases. And it’s not something one ever truly masters. You’re always learning, adapting to some new trend or security feature, and uncovering new methods for solving cases. You’re on the frontlines in combating the $1 trillion a year industry known as fraud. And rarely are two days alike.

Your coworkers have no idea what it is you do. The certificates you get are often the same ones military and law enforcement receives. And you’re even eligible to test for and receive a PI license! It’s a job that includes lots of tech, sometimes being on call for a court appearance, and is rewarding in and of itself. I mean, I get to work on some really interesting cases and see things few others can ever testify of. We see a lot, learn a lot, and the feeling you get when you crack a case…it’s amazing.

So whether it’s a dude claiming to be a vegetable who is using dummy LLCs in a real estate pyramid scheme, another dude who claims to be broke to sneak his way out of lawsuits while he liquids his assets, transferring them to Canada where he happens to be a multimillionaire, or it’s finding a mother pretending not to know the identity of the father so she can collect all the life insurance on her deceased toddler, the cases are never the same.

Eight years ago I didn’t know this position even existed. I was experienced in marketing, sales, writing, customer service, and project management. I started out in data entry, did some editing, and then stumbled into a super fun and wildly different day job. It’s perfect for the writer and mystery lover in me. And I can’t wait to take it to the next level.

Case Study – Data Doesn’t Lie

Well, data doesn’t lie unless the data is wrong. I was once delivered a task that seemingly no one else wanted: create quarterly reviews for our top fifty clients, plus a company-wide edition. At the time, the reviews were simply snippets cut out from a report auto-generated by the company’s online database and slapped into a PowerPoint. No one ever double-checked the data and not much thought was out into the presentation.

Personally, I feel PowerPoint is best used for presentations only. I found it restrictive for my purposes. And it didn’t take long for me to notice several inconsistencies in the data presented on the auto-generated report. By the way, this is a key skill for data analysts: the ability to recognize patterns and identity errors. If you enjoy puzzles and patterns, this is the job for you.

The first thing I focused on was getting access to the raw data in the form of Excel sheets. Once I properly organized and sorted the sheets, I went row by row until I was able to identify each and every error within the datasets. Mastering Excel is fundamental for data analytics.

For the first couple of years, I transitioned the final product for reviews from PowerPoint to Adobe PhotoShop. Of course, that was only a stopgap. After I learned the basics of Tableau, the goal now is to generate all quarterly reviews, and other reports, in Tableau itself. This requires me learning Python and R programming, which I am in the process of achieving. Albeit, slowly but surely.

My work on this project redefined the role of data, in general, and quarterly reviews, specifically, for not only the company, but our clients as well. It impacted marketing, sales, and internal review processes. I was able to generate specific charts and data that were used in large sales meetings, outside of just quarterly reviews. For example, I could generate weekly assignment turnaround times or chart usual, but vital statistical information.

We live in the Age of Information, and data is king. Three crucial duties a data analyst must fulfill: gather and interpret data, identify errors, and present the data in a visually pleasing manner. I know it sounds boring at first, but I love it. The raw data helps you get to the root cause of issues and can assist in helping to improve an organization from the inside out. I literally learned this job in the fly, with no assistance or guidance.

So the lesson is simple: never be afraid to take on new challenges. The risk of failure is worth discovering if it’s something you might love. It also reinforced my past experience in journalism with regard to fact checking. I am constantly “fact checking” the data I run to ensure accuracy. It’s crazy to think just how versatile the writing skill set really is…