Deep inside Palatine Analytics, an AI-powered people analytics software company, software engineers are at work on a secret project that envisions to create technology which will automate and utterly transform human resources decisions.
Indeed, Palatine Analytics can already predict employee performance four months into the future at a 75% success rate, which may eventually mean that a computer will be in charge of your career. Although the concept of being fired or promoted by a machine sounds scary, Palatine Analytics' founders claim that they are solving one of the biggest tragedies of the working world – the thousands of unfair, biased, and wrong-headed employee promotion and termination decisions made every single day. “It is incredibly sad to see people being promoted or fired unfairly," Palatine CEO, Archil Cheishvili said. "Further, these biased HR decisions have a catastrophic impact on companies’ survival."
Palatine Analytics' software incorporates data from a variety of sources but the main source is the feedback you get from your co-workers and supervisors. Employees at companies observe a colleague’s behavior and select the attribute they want to rate them on. A rating from 1 to 10 is entered into the platform, along with a brief description justifying the numerical rating. Then, the feedback is reviewed and digested, with insights about employees’ past, present, and future performance given as an output.
The software does three things. It collects, filters, and analyzes data. First, Palatine Analytics collects data through 360-degree employee feedback. Palatine Analytics tries to solve one of the biggest problems in the performance management industry: lack of data resulting from people not desiring to provide feedback. Through Palatine Analytics' software, people provide roughly 5 feedback points per week—which is, according to Archil, 500% more data than competing products.
“We maximize the number of data points by employing an automated system of rewards, reminders, and penalties. For instance, one of the strongest incentives for employees to provide feedback is the desire to view the feedback that you yourself have received. Under our system, employees receive notifications when they are evaluated but are barred from viewing said evaluations until they provide the required number of evaluations each week. This “curiosity” feature has been a very effective incentive because the last thing employees want is to know that people are talking about them yet be unable to view what's being said."
Second, the platform cuts through evaluation biases using AI and assigns each employee a score indicating how accurate of an evaluator he or she is. Therefore, in compiling evaluations, the software not only takes into account the raw scores employees receive but also weighs these evaluations according to the characteristics of individual evaluators. "For example, an employee who always gives 10 points to other employees, all else equal, will have a lower accuracy score because everybody cannot deserve perfect evaluations all the time," says Archil.
Third, Palatine Analytics' insights are powerful. “Our goal is to create a software that uses the AI’s understanding of not only an employee’s past and present but also their future via predictive analytics. This, in turn, grants actionable insights for employers about who should be promoted or fired and how teams should be assembled," Archil added. “We have developed a technology that, 75% of the time, correctly predicts whether your performance (quantified in a score) will be higher or lower in 4 months compared to your current score,” says Palatine Analytics' CEO Archil. “Using our predictive abilities, we empower managers to make smarter management decisions”.
"There has been so much innovation in many applications of AI, except the people analytics field. I believe that people analytics is the most important field compared to any other fields where AI can be applied because employees are the most valuable yet most difficult to quantify resource that companies have.”
Palatine Analytics wants day-to-day management—hiring, firing, decision-making—to be guided by software.
The team behind Palatine Analytics are grads from Harvard College and alumnus of Bridgewater Associates, who have started more than 5 companies and worked at Google, Facebook, and Goldman Sachs. According to Archil, Palatine Analytics is approaching a dozen major corporate clients, 20% month-over-month growth, and has received an undisclosed amount of seed funding from prominent angel investors and funds.
One day, he envisions a world where promotion and termination decisions are made accurately by a machine that is intelligent, fair and does not discriminate.