Yobs is a RESTful API-driven talent assessment service. The Yobs API uses resource-oriented URLs, supports HTTPS authentication and HTTPS verbs, and leverages JSON in all responses passed back to customers.
Yobs is used by customers in a wide variety of sectors, and supports a range of talent assessment insights. For a full list of our features, please see the Yobs AI Talent Assessment Report section below. Contrary to many cognitive services, we leverage a unique, human-in-the-loop approach to ensure quality and speed of service. For every media file passed to the API, we generate a confidence score based on the quality of the media file. If the confidence score is not high enough, the media file is sent to our team of trained reviewers who generate the report manually using our internal reviewer tools. We use a mix of human review, statistical checks and machine learning checks to ensure quality and speed.
This Product Guide is designed to help customers get up-and-running with Yobs’ AI talent assessment services, both by providing the necessary context to understand the talent assessment industry and its regulations, and by giving technical guidance on how to work with the Yobs API. Please note, you can find our technical API docs here. Before gaining a Yobs sandbox or production account, you must first schedule a call with a Yobs Customer Success representative to create and credential your account.
Click 'Request API Key' to get in touch with a representative.
Yobs follows a standardized assessment process:
Yobs’ API is flexible enough to support a range of workflows for integrating an AI talent assessment into your platform. At a high-level there are two options, each with unique pros and cons:
API clients can currently choose from the following list of features to have included in their Talent Assessment reports. Yobs outputs are grouped into the four types of insights they represent: Demographic Insights, Emotional State Insights, Thinking Style Insights, Big 5 Traits Insights. Yobs scores are all outputted as percentiles.
Definitions of the outputs are available in the table below.
I. Demographic Insights :
II. Emotional insights :
III. Communication and work style:
IV. “Big 5” Personality traits :
• Conscientiousness (Detail-oriented/Organized vs. Easy Going/Careless):
A tendency to be organized and dependable, show self-discipline, act dutifully, aim for achievement, and prefer planned rather than spontaneous behavior.
• Agreeableness ((Friendly/Compassionate vs. Challenging/Detached):
A tendency to be compassionate and cooperative rather than suspicious and antagonistic towards others. It is also a measure of one's trusting and helpful nature, and whether a person is generally well-tempered or not.
• Openness to learning (Inventive/Curious vs. Consistent/Cautious):
Openness reflects the degree of intellectual curiosity, creativity and a preference for novelty and variety a person has. It is also described as the extent to which a person is imaginative or independent, and depicts a personal preference for a variety of activities over a strict routine.
• Extraversion (Outgoing/Energetic vs. Solitary/Reserved):
Energy, positive emotions, surgency, assertiveness, sociability and the tendency to seek stimulation in the company of others, and talkativeness.
• Neuroticism (Sensitive/Nervous vs. Secure/Confident):
The tendency to experience unpleasant emotions easily, such as anger, anxiety, depression, and vulnerability. Neuroticism also refers to the degree of emotional stability and impulse control and is sometimes referred to by its low pole, "emotional stability”.
Now it's time to continue with the next steps to edit your website pages. Usually there are 2 types of content that will be edited, which are the following.
Here are tips on video recording to ensure the Yobs API returns an accurate report:
1. Each video URL length should be 90 seconds in aggregate minimum. For example, if a person recorded 3 video answers of 30 seconds each, you should combine the 3 videos in one video of 90 seconds in aggregate before passing it through the API. Shorter videos may lack enough data to offer compelling insights.
2. The following video codecs are supported: .MOV, .WMV, .FLV, .AVI, .MP4, .WEBM
3. Video should involve only one (1) person; the face should be clearly visible in the video. No full body. Face and shoulder view is best.
4. Audio should be of acceptable quality. Background disturbance like music or loud background noise may affect the quality of the analysis.
5. Video should be of acceptable quality. We recommend at least 360 kbps, but there is no need for HD or ultra HD quality. Low quality image may affect the quality of the analysis.
6. The person must be speaking in the video. If the person was silent for the duration of the video, the API will return an error.
7. The limit is 10 media files per POST request for a person. Exceeding this limit will create a 400 error saying "media exceeds limit". Clients can concatenate media files to circumvent this limit.
8. The person must be answering behavioral questions in order to maximize the accuracy of output. Examples can be provided by our customer success team, such as "Tell us of a time where you overcame a challenge with a team?"
Once your API account is credentialed and you are authenticated, your dedicated account manager at Yobs will work with you to review your implementation plan and ensure you are compliant and following best practices. We will also share inspirations, UI/UX assets, and product guidance to help you build the best possible experience for users whether they are customers, users or internal teams.
We will also share training resources to help you get familiar with the science behind Yobs talent assessment reports. For our enterprise customers, this can take the form of workshops and scheduled group sessions.
To learn more about the science behind the service, and how the service infers talent traits from voice, speech and video data, please visit our science page.
Yobs conducted a validation study with an independent group of I/O psychology researchers to understand the accuracy of the approach to inferring a personality profile for the automated portion of the service. Yobs worked with a team of leading Psychometrics researchers to measure the accuracy of the service on a sample of between 600 and 1200 participants for all characteristics. To establish ground truth, participants took two sets of standard psychometric tests:
Yobs then compared the scores that were derived by its models with the survey-based scores and expert observer scores for the video interview users. Based on these results, Yobs determined the average Mean Absolute Error and average correlation between the inferred and actual scores for the different categories of personality characteristics.
Based on the results of this validation study, the personality scores predicted by Yobs were shown to be strongly, positively correlated with all 'actual' personality scores. Yobs was also shown to outperform other popular cognitive services that were benchmarked in the study.
Review our API Docs here.