Employment Foresight has a number of uses for human resources and legal teams in businesses. The software uses machine learning to identify hidden patterns in judicial rulings, enabling users to navigate difficult areas of employment law and reach more informed decisions around issues such as reasonable notice, employee drug testing, worker classification and exemptions to overtime.
The platform collects and analyzes the facts and findings from thousands of previous employment cases, and uses the information to predict how a court might rule in new circumstances. This achieved through statistical methods based on machine learning, focusing in the identification of relationships between different factors like the industrial sector; length of employment; and employee’s position. This is to gain an insight into the industrial relations process.
Digital Journal: What are the challenges facing human resources professionals today?
Ben Alarie: The challenges facing HR professionals are essentially the problems that all organizations face. Technologies and industries are constantly changing, often forcing tough decisions for how organizations can best meet the needs of their employees and customers. On the employee front, organizations must address sensitive issues such as pay, discipline, performance, retention, and retraining, among others.
Compliance with legal requirements is critical, and non-compliance can be costly. For many organizations, solving these challenges cost-effectively and efficiently is itself a challenge. This is where software that helps to identify what might be needed to address a particular employment law issue is valuable.
DJ: How complex is employment case law today?
Alarie: There are many complex issues in employment law. The complexity is revealed by the thousands of cases that go to courts and tribunals across the country every year. If there were easy legal answers to these disputes, it would not make sense for parties to litigate as frequently. The silver lining to this complexity is that all of this litigation has generated a trove of data that we can use to train our algorithms. Having the decisions from these thousands of cases means that we can leverage them using machine learning to potentially reduce the amount of litigation in the future.
DJ: What types of cases or issues are most challenging for the workplace?
Alarie: Some of the most challenging questions in employment law relate to “line drawing,” such as: “given what has happened, would a court say that I had cause to dismiss this employee?”; “if we change the responsibilities associated with this position in this way, would a court characterize those changes as a constructive dismissal?”; “in these circumstances, would it be legally defensible to insist on a drug or alcohol test for this worker?”; “given the kind of job that this person held, their level of education, years of employment, etc., how much reasonable notice would a court say I would have to give them?”; and so on.
These are the kinds of questions that Employment Foresight addresses.
DJ: What solutions can Employment Foresight provide?
Alarie: Employment Foresight is a data-driven court outcome predictor. It collects and analyzes the facts and findings from thousands of previous cases to predict how a court would rule in new circumstances. It uses sophisticated statistical methods based on machine learning to reflect how judges have weighed various aspects throughout the body of case law. Employment Foresight allows for more accurate, precise and comprehensive legal research than existing tools.
Its approach based on machine learning empowers the user to consider all of the relevant factors and all the existing case law at once, which even the best lawyers cannot possibly do in the available time.
DJ: How did you develop Employment Foresight and what were the main challenges?
Alarie: Our company was founded in the spring of 2015 by law professors and machine learning specialists. Through many months of effort, we used this combined expertise to build and bring our first product, Tax Foresight, to market. The launch of our second product, Employment Foresight, is exciting because it represents the application of this accumulated learning to a new area of law.
We have come a long way since 2015, and we have relied on our expanded team (now a total of 18 legal researchers and software developers) to help us develop this latest product. There has been the equivalent of more than 25 years of full-time human effort invested in building Employment Foresight.
As in any startup, one of our challenges is effectively managing the company’s changes as it grows and its various functions develop and specialize. To harness the skills of everyone on the team, we have had to develop appropriate communication and internal processes. In some ways things are easier to get done when there are just a handful of people working together.
In other ways, of course, things are easier with a bigger team; people can specialize on the sorts of tasks and responsibilities that they find most rewarding for them. Having a law professor do payroll is not ideal!
DJ: How does the machine learning component learn?
Alarie: Machine learning involves the use of sophisticated statistical methods to analyze the factors that impact court decisions in any given area of law. Our legal research team identifies and maps the important factors in a given area of law and carefully processes the full-text of all the cases in accordance with those factors. This is an academic and labor-intensive process that allows us to turn the unstructured data of court decisions into structured data with which to train the system.
DJ: How have you tested out the platform?
Alarie: We have a strong network in the legal community as a byproduct of our founders’ positions as law professors at the University of Toronto. Our product has been tested by employment lawyers in large Bay Street law firms, in-house lawyers at large corporations, and HR professionals, and the feedback has been excellent.
DJ: What has been the response from human resources professionals?
Alarie: HR professionals are impressed with the ease of use and practicality of the software. They have noted that they deal with these kinds of issues on a daily basis, and Employment Foresight would allow them to make better decisions more confidently.
DJ: Which types of companies or sectors have adopted the system?
Alarie: Part of the benefit of Employment Foresight is that it is applicable to every industry, as it considers the entire body of case law on each employment question it addresses. We just launched this week and we already have paying customers. We predict that the early adopters will be tech-savvy individuals working in visionary organizations excited to use the newest tools to empower their teams with the most accurate, precise, and comprehensive legal information.
DJ: What other projects are you working on?
Alarie: As we continue to learn, we actively assess additional areas of law that may be suitable for the application of machine learning. Our methods are also applicable in other geographic areas, so we are expanding to other jurisdictions. We have a dynamic product development team with members who are excited about the next adventure, whatever that may be. In addition, the quality of Legal Foresight depends on keeping our current products up-to-date.
We are constantly adding new cases to the system to train the machine learning algorithms on the latest cases. Moreover, our users are constantly identifying legal questions that they regularly confront, so we continue to expand and improve on the coverage of our products.