Take a Deeper Dive

Our Technology

Data transformation begins with our grammar and knowledge engineering tools

  • Cleans the data
  • Tags every word and phrase
  • Identifies syntax, semantics and thematics
  • Extracts events configured to company needs
  • Dashboard displays for convenient usage

Key data is included in the decision making process

  • Faster and more complete access
  • New knowledge can be acquired
  • Consistent measurement of metrics
  • Additional data for warehouse mining

Important data discovery for underwriting and claims

  • Events that happened, did not happen, may happen
  • Identify new exposures and concentrations of emerging risks
  • Assistance with key operational barriers where staff are constrained by reading information
SD Refinery-technology-complete

Simply Stated
For operational executives and those that want to maximize the profitability and productivity of their business

Can a Computer Read Faster Than a Person?
Reasons that an executive should care about sentence data

What is Sentence Data?
Sentences, sentences and more sentences. Understand what constitutes sentence data and its applicability to SD Refinery

Using Technology to Read and Interpret Sentence Data
The technology that makes it possible to read and interpret sentence data

Applications in Risk and Insurance
Sentence data is everywhere
in the insurance industry waiting to be utilized

Applications in Healthcare
For an executive in healthcare, the advantages of using technology to process sentence data are enormous

A simplified view of SD Refinery's technology

Preparation is the Hardest Part

Sentences are written by people that are usually in a hurry. As such they usually contain a lot of typos, acronyms and miss-spellings. A critical step in the refining process is preparing the data so it can be effectively processed

Breaking Apart the Sentences
Documentation written in a professional setting - whether it be an underwriter, claim adjuster, or surgeon - may vary from complete sentences to short snippets of information. Regardless of the form, grammar rules are applied through a technology to break apart each sentence and to understand the context of the information

Assembling the Pieces
Critical facts extracted from sentences can be used in any manner desired by a customer. The pieces of information can be summarized into a dashboard or loaded into database fields used by predictive models. Additionally, the pieces are perfect for developing the training sets used in machine learning

Can a Computer Read Faster Than a Person?

Yes, computers can read faster than a person!

A person reads about 13 sentences per minute

SD Refinery technology reads over 250,000 sentences per minute

Most business systems use coded data to generate reports and analyses. Data is coded by people reading sentences assigning codes to it

Coding Data is Obsolete
Historical way of accessing the knowledge in sentence data is by assigning codes to it:

  • Slow
  • Inefficient
  • Ineffective
  • Archaic
  • Expensive

Use SD Refinery as a Better Way

SD Refinery's technology can read and interpret sentence data according to the unique needs of a client. This is the efficient and effective alternative to people coding data

What Is Sentence Data?

The Basics of Sentence Data in Businesses

  • Written documentation = sentence data
  • Sources: submissions, reports, forms, note diaries, letters and emails
  • Wherever employees are typing sentences, sentence data is being created
  • Sentence Data in Relation to Business Data
    Businesses capture more data every day from different sources including images, video, web, email, and social media. Regardless of the source, sentences written by employees capture the meaning and relevance of the data

    What is Sentence Data? Here's a 1 minute video with an explanation!

    Using Technology to Read and Interpret Sentence Data

    SdRefinary-breakthroughs

    There are many new tools and strategies for working with business data: business intelligence, data mining, data analytics, predictive modeling and machine learning

    SD Refinery's technology makes every tool better by identifying and extracting critical events and activities that happened, did not happen or may happen

    Business executives using sentence data - rather than coded data - were able to achieve these breakthroughs:

    • 33% decrease in expense and 66% decrease in time to perform contract analysis.
    • 93% reduction in defective product identification cycle time.
    • 200% increase in the number of fraudulent claims identified.
    • 300% more cost identified when shoulder injuries occur during unloading.
    • 400% more insured accounts identified with bridge exposure.

    Applications in Risk and Insurance

    "Risk and Judgment ... Always Hard To Measure"
    Productive Economy = Constant Improvement + Rapid Rate of Change
    "What is good for the economy is not so good for the measurement of risk"

    Underwriting
    If an underwriter could see the future, then the perfect policy and premium could be written for an insured. But then why wouldn’t the insured simply hire the underwriter so they could avoid the loss before it happened?
    Two constant realities:

    • The past losses will always be the best indicator of the risk that lies ahead
    • Written descriptions of risk are far more informative than structured, coded data

    SD Refinery’s sentence data refining process creates new, better and faster ways to measure risk and uncertainty by using technology to read written descriptions of risk. This leads to increased revenue and decreased expenses

    Claims
    Claim adjusters create a wealth of knowledge by documenting in claim notes the decisions they make about each and every claim they handle. Typically, claim notes are largely ignored except for periodic audits or exams that may or may not identify issues. Using technology to read and interpret every sentence in every note in every claim file, creates a paradigm shift in the way to tap the knowledge repository existing in claim notes 

    • Read every sentence written
    • Create multiple operating metrics
    • Identify issues and opportunities

    These metrics increase profitability by generating savings on claims and lower operating expenses

    Applications in Healthcare

    The Subjectivity of Healthcare
    Every human body: same injury and same treatment = widely different outcome

    Healthcare professionals write sentences to document their judgment about relevant factors impacting outcomes

    Healthcare systems use coded information to record information and make decisions. Meanwhile, written sentence data is a wealthy knowledge repository largely ignored because conventional systems are unable to use sentence data

    Surgical Operations
    Pre-operative and post-operative reports written by surgeons contain extensive documentation about the profile of the patients, extenuating factors, techniques used to perform procedures and medical devices and equipment used throughout the surgery

    SD Refinery structures the sentence data and makes it possible to improve clinical outcomes by more quickly and efficiently identifying both good and bad techniques

    Medical Products and Devices
    Innovation in medical and surgical equipment continues at a very rapid pace. While this is the desired long-term impact, the short-term can create problems because products, devices and equipment can fail or be used incorrectly.

    If a defective product or technique were known ahead of time, the perfect system could be created to identify when and where it was happening. Because it isn't known until it happens, risk exists.

    For one product manufacturer, a 93% improvement (reduction) in the cycle time to identify defective products was achieved by using sentence data rather than coded data.