Skip links

oseberg’s data odyssey continued (part 3): unstructured data is a party foul

Structure this

Think of structured data as data in spreadsheets and unstructured data as data in images like PDF of TIFF that would need to become structured before you could do anything semi-awesome with it. If structured data is the darling of meticulous spreadsheet enthusiasts, unstructured data is the unruly party crasher that refuses to conform to our rigid expectations. 

Structured data

Structured data is quantitative data in the form of numbers and values.

Unstructured data

Unstructured data is qualitative data in the form of text files, audio files, and video files.

The analysis would then be drawing on large structured datasets to identify patterns in order to make decisions interplaying with the belief that large data sets offer a higher form of intelligence and knowledge and can generate previously impossible insights, with the aura of truth, objectivity, and accuracy as if we believe that within the zeroes and ones lies the answer to life, the universe, and everything else. The reality is that humans are just not able to process too many data points at once. 

Fungi of Innovation or Just Spreadsheet Junkies?

Data companies sprout like mushrooms after a rainy day, and they emerged in the oil & gas space in spades in the 90s to help companies lower lease operating expenses and increase reserve growth – armed with the next iteration of the spreadsheet. So 70% of the global oil & gas software market addressed production and operations functions. Unconventional operators drove growth in demand for comprehensive “ERP-like” systems to help manage substantive amounts of data and operations associated with high-intensity drilling. 

In parallel, the big data and business analytics market would become a $200B market. Even so, there was a staggering amount of unstructured data needed to drive a systematic process trapped in paper, .pdf, and/or .tiff files (images). Well spacing (well-well communication), lowering LOE, and reserve growth remained top of mind for industry executives – but to just me, it seemed, it was a little like putting the cart before the proverbial horse – what about the documents that actually underpin and govern all these reserves? Folks were clinging to these ancient relics like physical land files, well files, mud logs, check stubs, regulatory filings etc. It wasn’t quaint. It was perplexing. 

Who Needs Modern Tools When You Have Dusty Archives?

Energy companies seemed stuck in data transformation hell- wrangling the structured data. So while they were busy standardizing, normalizing, cleansing, the structured data of the 70s, I looked beyond that horizon and set out to build infrastructure and technology that was going to create structured data from unstructured documents.  

We created Oseberg to be the first to create structured data from unstructured documents at scale. But we needed a LOT of capital for product development. 

(let me grab a kombucha and break here for a minute)

hanging by a thread,