Along with a fleet of drivers and passengers armed with smartphones, Uber relies on cutting-edge technology to manage its ride-hailing service. The company uses artificial intelligence and machine learning powered by complex algorithms to connect passengers with nearby drivers.
In order to keep these systems running and scalable for the future, Uber needs new IT staffers. To fill its empty tech seats, the ride-sharing company is turning to third-party and custom-built systems to find and vet the right data scientists, designers, project managers and engineers.
To find new recruits, Uber uses HackerRank, a technology job recruitment and job listing site that aims to match coders and engineers with high-tech firms. HackerRank’s two-way video application CodePair, allows Uber to watch a job candidate write code and solve problems in real time. Uber also uses Google Docs and Skype for sharing files and conducting interview sessions, respectively.
Once it identifies the recruits it wants to hire, Uber deploys two of its own tools, the Interview Question Database and the Candidate Relationship Tool, to further weed out candidates.
The Interview Question Database is an interactive technology questionnaire, which consists of algorithmic computing problems that are submitted and ranked in importance by current Uber engineers and software coders. Uber candidates are asked to fill out the online form. Uber technologists regularly update the questions and managers rate them by difficulty, the level of the applicant’s understanding, experience, and the requirements of the position the firm is attempting to fill.
“The engineers are able to vote certain questions up or down on the effectiveness and the usefulness of those questions,” says Greg Campbell, who recently joined Uber as director of technology recruiting with a mandate to fill Uber’s New York and Seattle technology and development offices. He adds that some questions are ranked higher than others in order to make sure that we're constantly fine-tuning the quality of those challenges. He declined to specify how many new technologist Uber is looking to hire.
A candidate takes Interview Question Database test on a laptop in the Uber offices in order to emulate a typical development environment. Uber does not place a time limit on the tests. Instead, the hiring team is looking for how the candidate processes and thinks through the often algorithmic-based problems.
“It's really more about what’s the algorithm in your computational thinking to get to a solution,” Campbell says. “And even if you didn’t get to a solution, there are all kinds of value in how you processed and systematically thought through the challenge.”
Uber also interviews and tests for emotional intelligence during the hiring process using its Candidate Relationship Tool. The system gives candidate real-world project scenarios to determine how they approach a problem, how they communicate and how they involve different members of a team.
“The collaboration piece is something we index when we are assessing and evaluating candidates: How they collaborated with small teams, one-to-few, one-to-many and how they brought people on board and got buy-in for an idea,” Campbell says. He adds that Uber looks for how candidates demonstrate flexibility when they come across a new or better project approach.
“We really try to think about those dynamics,” Campbell says. “Not just in a vacuum but how have they done this with limited resources? How long have they had to undo things or unlearn things in order to get to a better place?”