NLP powered “LessLer” helps organizations replace lost experience and knowledge

As of February 2017, job cuts resulted in a loss of over 440,000 positions in the global oil and gas industry according to Houston-based consulting firm Graves & Co. With these job losses, knowledge is lost and so is valuable mentorship for less experienced staff. This exposes your company to increased risk in repeating mistakes of the past.

Understanding this dilemma Waterford Energy Services set out in 2016 to invest in the development of a solution utilizing Artificial Intelligence to help address this significant industry problem. A team of data scientists and software development experts were hired to form Waterford’s Data Science Team. After several years of development, the result is a new software tool called “LessLer”. LessLer allows individuals to quickly access unstructured documentation in a mentored environment to solve their problems, answer their questions and ensure that past lessons are indeed learned, not lost.

With industry changes in staff demographics, forced retirement, layoffs, and a transient workforce, we now have a new era of business from which to work. Teams with lower experience levels have to plan and execute highly complex projects while doing it on time and on budget without harm to the environment. These teams must educate themselves on past practices and processes as a fundamental stepping stone in their success.

Blair MacDougall, President of Waterford Energy Services says, “With the recent downturn in the industry, key personnel have left at an alarming rate taking their tribal knowledge with them. Most of this knowledge is documented in massive amounts of unstructured data which, in our opinion, is too difficult to utilize effectively. It gets lost on shared drives, lost when staff leave, too difficult to access, and so on. We are curbing the tribal knowledge loss by providing a tool that ingests all of that documentation and presents it to the tech-savvy, yet less experienced teams that appear to be more prevalent today.” MacDougall continues, “We’ve also listened to our customers who want the ability to have Subject Matter Experts and mentorship functionality within the LessLer platform.”

Lessler is software under development in Waterford’s offices that is a step change in how unstructured data is utilized in the industry. It is a software tool that organizes virtually all documents and transforms it into actionable knowledge. Mike Cullymore, Senior Software Specialist with Waterford Energy Services, says, “Think about all the data we acquire in a day, let alone a year. It’s massive intellectual property that drives a company forward. LessLer is designed to harness the company’s history in the form of emails, documents, reports, equipment manuals, product specifications and incident reports.”

As a mentoring tool, Lessler also permits oversight by subject matter experts, or other stakeholders such as regulatory agencies and service companies. This enhances the learning experience of users and ensures that key decisions using LessLer are peer reviewed.

Lessler is intuitive using data visualization techniques to create a vibrant user experience. Data Scientist Mateus Pereira, a native of Brazil, applies his specialist skills in data visualization and ranking initiatives. He says, “LessLer presents data in a rich graphical interface that simplifies the user experience to find answers to complex problems. The user experience is currently being perfected so unwanted responses can be rejected and acceptable responses promoted, allowing LessLer to learn from the user.”

Overseeing the overall construction of LessLer is Lead Data Scientist Elham Etemad, PhD. Her focus is the integration of two main components: A.I. and Mateus Pereira’s data visualization. LessLer uses natural language processing (NLP) techniques to find candidate documents for solving the user’s problem. This component consists of query processing and document retrieval modules. The query processing module applies query correction and suggestion to help the users in improving their questions’ precision. The document retrieval module finds the most relevant documents and sort them based on their textual similarities with the question and feedback from users.

What is known for certain is that the world is changing and it’s changing quickly. Within Exploration and Production companies, there are unprecedented amounts of documentation that is highly under-utilized. Companies who will lead the industry are those that adapt, learn, and implement tools to use their data effectively, creating value through effective decision making. The next generation of energy workers are in need of better ways to unlock their historical data to ensure high risk decisions are informed decisions and LessLer is being developed to do just that.

Waterford Energy Services Inc.- LessLer Team Developers: (From Left to Right) Blair MacDougall (President), Manav Sharma (Data Scientist), Raghav Gupta (Developer), Navkaran Kumar (Developer), Elham Etemad, PhD (Sr. Data Scientist), Mateus Pereira (Data Scientist), Kalene Hines (Project Engineer), Mike Cullymore (Sr. Software Specialist).