Presented by Indeed

If we look back to only a decade ago the best jobs were often word of mouth, and even when published, were difficult for a job seeker to find. Since then, aggregating jobs from thousands of sources has resulted in everyone online having equal access to job information and has fundamentally changed how people search for work. But with most jobs now online, finding the best-fit jobs without the help of artificial intelligence would be like finding a needle in a haystack.

AI powered by job seeker behavior data

At Indeed, the core of our product offerings is a search engine to help job seekers easily find relevant jobs. When one of our 200 million monthly users searches for a job title and location in one of the 60+ countries we serve worldwide, our algorithms use fourteen years of job seeker behavior data to learn and continuously improve the relevance of the jobs that are displayed. The engine uses AI to rank matching jobs and help the job seeker understand when a job is a great fit. In the last year alone, we were able to double job matching relevance.

One of the earliest examples of AI use at Indeed was salary prediction.  Every job seeker cares about salary, yet many companies do not provide salary information in their job descriptions.  By using third-party resources and hundred of millions of salary points from historical job search data, Indeed’s algorithms are able to predict a salary range for jobs when a company does not.  Employers were not always thrilled, but it was the right thing for job seekers so we quickly allowed employers to override, and thus help train, the salary predictions.

AI powered by search behavior data is also used in many other ways, such as predicting jobs that have reasonable commute times.  By monitoring where job seekers look at jobs over thousands of similar searches, patterns evolve and train the algorithms to offer jobs that have reasonable commutes – all without the algorithm knowing anything about the local geography, train schedules or traffic congestion patterns.

New AI technology to improve the job seeker experience

Using job seeker behavior and outside databases was a great start at powering AI to increase relevance. With over 90 million resumes, billions of job descriptions, and more reviews of employers by employees than any other site, there is also a wealth of unstructured data available to Indeed to mine which increases the accuracy of our matching algorithms. Job descriptions are explored with Natural Language Processing algorithms to extract the details of experience requirements and soft skills that a company prefers. For job seekers, NLP is used to better understand personal desires from resumes and company reviews. This data can significantly increase the prediction success of a match and paint a deeper picture of an opportunity.

Moving beyond resumes and job descriptions

Job descriptions and resumes continue to be the first impressions available to job seekers and employers and are still the standard inputs for matching algorithms. However, these documents are one-dimensional views of the workplace fraught with the potential for unintentional human biases.

With today’s job descriptions, it is often hard for a job seeker to get a good read on whether they are a good fit.  Worse yet, commonly used language can discourage women, older workers and other groups from applying for jobs when otherwise they would be a great fit.  With resumes, it is difficult to reflect the depth of experience of a job seeker.  A resume written to target one employer is likely all wrong for another. Indeed is working on products to improve these prehistoric documents. An example is AI-driven assessment technology.

When looking for employees with specific skills, AI can search the job description and suggest assessments that would be appropriate for the job. It can further reduce the time needed to understand qualifications by automating the testing, scoring the results, and recommending candidates.  By allowing a job seeker to interactively provide work samples that match an employer’s requirements, assessments provide employer confidence in the prospective employee. Executed well, these assessments can help remove the human biases that result in diverse-challenged organizations.

For the job seeker, knowing more about what the employer needs and that their skills are a match can be very empowering. Assessments can be recommended based on the experience found in resumes. These assessments can provide the job seeker with advice about skills needing improvement, jobs to apply for, and certainty in understanding how they rank in targeted skills compared to others applying for the same jobs.

AI will continue to reshape the hiring process

AI is rapidly changing the recruiting industry. Traditional job-hunting resources — resumes, networking with friends, industry associations – are always going to be helpful in finding a great job.  But they are a small slice of the universe of what can be done today.  When so many jobs can be viewed online with ever-evolving AI-powered matching technology — where jobs come to you — these traditional methods must be augmented with online search, assessments and other innovations to find the best job available and make that dream job come true.

Raj Mukherjee is SVP of Product at Indeed.

Sponsored posts are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. Content produced by our editorial team is never influenced by advertisers or sponsors in any way. For more information, contact