Palantir – Big Data Possibly Helped Catch Bin Laden
Palantir Technologies, a Palo Alto headquartered big data analytics company, is rumored to have possibly helped in tracking down Osama bin Laden . Although no concrete sources have confirmed this rumor, it is well known that one of Palantir’s main products, called Gotham, is used by counterterrorism analysts at numerous US government agencies. They sold almost exclusively to the government sector until a few years ago. Their reputation is shrouded in mysticism due to their roots in spycraft, antiterrorism, and intelligence collection. However, it is undeniable that Palantir’s products have had significant impact cutting through the noise of big data to draw definitive connections across disparate dimensions.
Palantir was cofounded by Peter Thiel, a billionaire most notable for cofounding PayPal. He supposedly named the company after the “seeing stone” from a J.R.R. Tolkien novel. Although they initially had difficulty getting funding, they ultimately received a $2 million seed investment from an unlikely source, the CIA’s venture capital arm called In-Q-Tel . Supposedly, Sequoia Capital’s Michael Moritz doodled on a notepad during Palantir’s funding pitch. Palantir is now one of silicon valley’s biggest private “unicorns” valued at over $20 billion.
Palantir offers two main products; Gotham and Metropolis, named appropriately after the cities in which Batman and Superman fought crime. Similarly, Palantir’s mission is to leverage their data manipulating acumen to change the world by connecting the dots. Gotham is used for various purposes including counterterrorism, whereas Metropolis is used mainly by financial service companies.
This product was formerly called Palantir Government and apparently uses a more focused approach to intelligence data analytics within the confines of privacy and civil liberties protections. It integrates structured and unstructured data and differs from notorious data mining techniques that the NSA has been accused of. The data storage container for Gotham is called AtlasDB and it combines the simplicity of NoSQL stores with the consistency of SQL databases. The product starts from multiple data sources and aims to translate it into dynamic models digestible through human-driven analysis. The main interfaces for visualizing data objects are graphs, geospatial maps, and object explorers, a browser for filtering data. The ultimate goal of the product is to “integrate, manage, secure, and analyze enterprise data.”
This product, targeted at the financial sector, utilizes a slightly different approach to understanding data. While Gotham relied on links between data objects, Metropolis takes on a more dynamic mathematical analysis of model behavior relative to time.  Palantir’s product description for Metropolis is to “integrate, enrich, model, and analyze any kind of quantitative data.”  The foundation of Metropolis is packaging data into computational models using a diverse library of pre-created statistical and mathematical operators. This modular approach to quantitative analysis allows for complex operations to be quickly tested by analysts and iterated upon. A comprehensive suite of visualization interfaces rounds out this product for use in providing usable and actionable results.
Although Palantir was initially associated with intrusive data collection, they have pivoted into being a reputable company with better intentions. Their products are now commonly used to detect fraud or aid charity organizations. In one example, Palantir was going to investigate illegal financial transactions connected to global human trafficking. In another example of Palantir’s morality, they turned down a large contract with a large tobacco company fearful that the data would reveal demographics most susceptible to cigarette sales. It appears that Palantir’s mission and long term vision have clear directive. As they expand to various other sectors, it will be interesting to see whether their convictions hold true in carrying out their mission of harnessing big data for good, similar to the super heroes that they’ve named their products after.