FAQ

Here are the main steps to get started with UltraHire:

1. Create an account. You'll need an email address and password to sign up for an UltraHire account. This will give you access to the recruiter dashboard.
2. Create a job posting. In the recruiter dashboard, click "Add new job" to fill out the details of the job you want to post. Include the job title, description, requirements, responsibilities, and any other important details.
3. Optional: Create a questionnaire. You can create a custom questionnaire with multiple choice or short answer questions for candidates to fill out as part of their application. This helps you gather structured information from candidates.
4. Coming Soon: Share the job. Once your job posting is live, share the link via your preferred channels like LinkedIn, Indeed, Monster, or your company career page. You can also share directly with candidates from your favorite candidates list.
5. Review candidate videos. Candidates will apply by recording and submitting a short video. You can browse, sort, and filter all incoming candidate videos from the "Applications" page in the recruiter dashboard.
6. Invite your favorites. As you review videos, add any candidates you like to your "Favorites" list with one click. You can then invite them to apply to other relevant jobs.
7. Request contact info. For candidates you want to move forward with in the hiring process, you can request their full contact information and resume.
8. Interview and make an offer. Coordinate next steps like phone screens and in-person interviews through the platform or directly with the candidates.

That's the basic process to utilizing UltraHire's video hiring platform!

Here are the key points regarding the AI technology powering UltraHire:

• Coming Soon: Video and text analysis - UltraHire uses AI and machine learning to analyze the video applications submitted by candidates. It can identify traits like enthusiasm, communication skills, personality, competence and culture fit based on facial expressions, body language, speech patterns and vocabulary.
• Coming Soon: Resume parsing - UltraHire can intelligently parse the resumes uploaded by candidates to extract skills, experiences, accomplishments and other key information in a structured format. This information can then be used to match candidates to job requirements.
• Coming Soon: Natural language processing - UltraHire employs NLP technology to analyze the answers candidates provide to open-ended interview questions. This allows the system to gain a deeper understanding of candidates beyond what is explicitly stated.
• Coming Soon: Recommendation engine - An AI-based recommendation engine matches the top candidates to relevant job openings based on their identified skills, video analysis and questionnaire responses. This helps reduce the hiring manager's screening workload.
• Coming Soon: Automated screening - For some roles, UltraHire can perform an initial screening of candidates based on pre-defined skills, experience and other factors to identify the top matches for human review. This automation further reduces recruitment cycles.
• Coming Soon: Personalized job recommendations - The system can recommend new job openings to registered candidates based on an analysis of their profiles, interests and previous applications. This improves the candidate experience on the platform.

In summary, the key technology differentiators powering UltraHire are its abilities to:

1. Analyze video applications beyond text to identify great candidates
2. Augment human screening and decision-making using AI and machine learning
3. Automate portions of the recruitment workflow to drastically cut down hiring cycles

The AI and machine learning algorithms are continually trained on more data to improve their accuracy and evolve with the needs of users.

Here are the keyways UltraHire ensures the accuracy of its AI-powered video analysis:

• It collects a large and diverse dataset of candidate videos and the "ground truth" traits identified by human raters for those candidates. This helps train the initial machine learning models.
• The models are continually re-trained as more candidate videos and human ratings are collected. This ongoing training improves the accuracy of the models over time.
• UltraHire employs multiple human raters to evaluate each candidate video and extract the displayed traits. It then takes the average of these ratings as the ground truth. This helps reduce individual rater biases.
• The human raters undergo stringent calibration training to ensure consistency in their evaluations. Their ratings are also periodically checked for accuracy. Inaccurate raters are retrained or removed from the system.
• The AI models make probability-based predictions instead of definitive declarations. For example, instead of saying "this candidate is enthusiastic", it may say "there is an 80% chance this candidate is enthusiastic based on the video." This leaves room for human judgment.
• UltraHire discloses the accuracy rates of its different models based on validation data. This allows enterprises to gauge how much to trust the AI predictions.
• For critical hiring decisions, human recruiters are still required to thoroughly review candidate videos to confirm or override the AI analysis. The system acts as a recommendation and screening tool, not a replacement for human judgment.
• UltraHire continuously monitors the performance of its AI models in production and makes adjustments where needed. It also solicits feedback from enterprises and candidates to further refine the models.

In summary, a mix of simulated and real-world training data, multiple human raters, stringent evaluation protocols, probability-based predictions, transparency around accuracy rates and the potential for human oversight all contribute to UltraHire's efforts to ensure its video analysis technology is correct and fair.