Intended purpose

Spark Hire AI helps teams make faster, more confident decisions without sacrificing quality. By surfacing consistent, explainable insights from One-Way Video Interviews, teams can identify top candidates sooner and stay competitive in fast-moving hiring cycles. 

Spark Hire AI includes three features that work together to support high-quality One-Way Video Interview review:

• Transcriptions generate searchable, time-synced transcripts to search or reference specific details
• AI Summary provides a quick overview of each answer to decide what to watch
• AI Video Review scores responses based on key traits to help prioritize candidates

These tools are designed to enhance—not replace—human decision-making. Hiring teams are encouraged to use AI insights as a starting point for further review and discussion.

Spark Hire’s approach to AI

At Spark Hire, we leverage AI to enhance the hiring process and support, not replace, human oversight in decision-making. In designing and implementing AI feature we adhere to the following principles:

• Fairness and non-discrimination: Spark Hire AI tools are designed to reduce bias by only taking into account objective data and criteria. To ensure fairness, we also conduct bias audits, confirming that our AI does not introduce bias from past screening data or human-made decisions.
• Customer-centric approach: Spark Hire builds tools to empower recruiters and managers, ensuring AI outputs are explainable, transparent, and easy to use, with continuous feedback loops for improvement.
• Human oversight: Spark Hire AI tools support recruiters and managers by providing insights, but the final hiring decisions are always made by humans.
• Diverse talent: AI tools help remove subjective barriers, giving more candidates from diverse backgrounds a fair opportunity to showcase their skills. Learn more about how we ensure fairness in our candidate selection process.
• Privacy and security: We take privacy seriously, implementing strong security measures to protect all data processed by our AI tools. Spark Hire’s AI tools are built within the platform’s existing security framework, which is designed to safeguard both customer and candidate data. This includes role-based access, secure passwords, and single sign-on (SSO); encryption of data at rest and in transit; and hosting on AWS with strong cloud security and continuous monitoring. The platform also meets standards like SOC 2 and uses redundant systems to support availability and reliability.

How it works

Spark Hire AI for One-Way Video Interviews includes three integrated features that work together to help teams review candidate submissions more efficiently:

1. After a candidate completes their One-Way Video Interview, Spark Hire’s AI Transcription feature automatically detects the spoken language and generates a synced transcript. This transcript appears under the video player and scrolls with the playback.
   • You can search, edit, or download the transcript
   • Clicking a line of text jumps to that part of the video
   • Closed captions are available during playback
   • Supports 50+ languages
2. Once transcribed, Spark Hire’s AI Summary feature reviews each response and generates a short summary highlighting the key points. These appear directly below each recorded response.
   • Helps you decide which videos to watch
   • Included in email notifications so reviewers can triage quickly
   • Summaries can be shown or hidden in the platform
3. When enabled, Spark Hire’s AI Video Review evaluates responses in English and scores them across 4 competencies and 2 personal qualities. Then, it provides the candidate with an overall score, shown in the Assessments list and the candidate’s profile.
   • Competencies: Communication, Execution, Comprehension, English proficiency
   • Personal qualities: Enthusiasm, Motivation
4. Each factor is rated as Strong, Neutral, or Weak, with rationale shown for each. The AI then generates an overall score (Strong / Moderate / Weak) to help prioritize candidates.

These features are disabled by default and can be enabled from the Job settings page, providing flexibility to manage the feature as needed.

You can submit feedback on AI scores to help improve the system. Together, these tools provide quick, structured insights that support better collaboration and reduce time-to-hire.

Accuracy & limitations

The AI Video Review feature was built through a process that promotes scientific rigour to reach a high level of accuracy.

Thorough testing

The AI Video Review feature was carefully developed to be accurate while protecting privacy. It was tested against human reviewers and showed strong alignment in scoring candidates’ responses. The AI reliably matches human judgment across all factors. Testing with over 68,000 interviews confirmed that AI scores correlate well with real hiring decisions, helping identify candidates likely to succeed.

Fairness

At Spark Hire, we developed our AI tools with fairness in mind, aiming to reduce bias throughout the hiring process. For more information about AI Principles at Spark Hire, refer to this page.

AI for One-Way Video Interviews is designed to focus on objective content in candidate responses. The evaluation process was built to exclude Personal Identifying Information (PII) such as name and location.

To help ensure our AI systems remain fair and aligned with best practices, we regularly conduct bias audits. These audits evaluate adverse impact across demographic groups, including gender, ethnicity and intersectionality.

Our commitment to fairness includes continued evaluation, transparency, and improvement of AI features as hiring practices evolve.

Limitations

• Transcript fidelity: The model evaluates text transcriptions of spoken responses, which may contain errors or omissions that affect interpretation. While Spark Hire uses an industry-leading speech-to-text provider with one of the lowest word error rates and highest accuracy available, imperfections in transcription can influence how both AI and humans understand a candidate’s intent.
• Positive presentation bias: As job applicants, candidates are naturally motivated to present themselves in the best possible light. This may result in rehearsed or generalized responses that obscure deeper behavioral signals, potentially limiting the depth of both human and AI scoring.
• AI interpretation boundaries: While the AI model follows consistent behavioral definitions, it may still miss subtle context, tone, or culturally nuanced cues that humans would naturally pick up. Final decisions should always involve human oversight.
• Limited factor coverage: The AI focuses on a specific set of competencies and personal qualities. It doesn’t account for all the job-specific traits, qualifications, or context a recruiter or hiring manager might consider when making a hiring decision.
• Supported languages: Transcripts and AI Summary support 50+ spoken languages, see the full list. AI Video Review only supports English.

Human oversight

While Spark Hire’s AI features are designed to enhance hiring workflows through automation, structure, and speed, they do not replace human judgment. All hiring decisions must be made by the recruiting team.

1. AI outputs—such as behavioral scores and response summaries—help surface patterns and provide additional context for evaluation. These models assess responses based on defined factors, but some qualities essential to hiring decisions—such as team fit, communication style, or other subjective traits—may not be fully captured. That’s why AI-generated insights should support, not replace, human review.
2. If a customer disagrees with an AI-generated output, they are encouraged to share feedback using the built-in tools available in the platform. This feedback helps Spark Hire refine and improve the models over time.

Transparency and explainability

Spark Hire provides clear and accessible information about how the AI tools work. Rankings in the app are accompanied by explanations and our resources explain how the AI features work.

For recruiters and managers

1. Recruiters and managers can view detailed outputs of Spark Hire AI features directly in the platform—such as interview summaries, behavioral evaluations, and response transcripts. AI Video Review scores include insight into the factors evaluated and how the model interpreted the candidate’s responses, providing transparency and enabling more informed decisions.
2. Spark Hire offers information about the AI tools, covering their functionality, benefits, and best practices in our Recruit Help Center and Meet Help Center. The documentation is regularly updated, and accessible to all users at any time, ensuring they have the support they need to effectively leverage our AI tools in the recruitment process.

For candidates

1. We offer candidates a dedicated article in our Help Center explaining how Spark Hire AI works. It can help them understand the process used to evaluate their applications and promotes transparency in the recruitment process.
2. Employers are responsible for determining whether they are required to notify candidates about the use of AI in their recruitment processes and provide any additional information, depending on the applicable laws and regulations in their jurisdiction.

Compliance

Spark Hire is committed to adhering to the laws and regulations that apply to Spark Hire as a vendor of recruiting solutions. We focus on developing and maintaining AI tools in line with legal requirements. However, as the employer, it is your responsibility to understand the specific laws and regulations that apply to your organization, including, without limitation, transparency requirements and to obtain consent. We strongly recommend that customers take the necessary steps to ensure their own compliance with applicable employment and data protection laws.

You, as the Spark Hire customer, must comply with any and all applicable laws and regulations (including, without limitation, the NYC Law 144’s requirements before activating and using an AI tools).

Share your feedback

Meet actively seeks feedback to refine our AI tools. Our commitment to continuous improvement means we aim to enhance our technology based on user experience and needs. You’re welcome to share your feedback with us in any of the following ways:

• In the product: Tap the thumbs-up/thumbs-down icon for AI Video Review ratings.
• In the chat: Reach out to our support team.
• Get in touch: Use this form to send your feedback to our team.