AWS Launches Amazon Connect Talent for AI-Led Hiring at Scale

Amazon Web Services launched Amazon Connect Talent on September 17, 2026, an AI hiring product for talent acquisition leaders managing scaled hiring, delivering AI-led interviews, data-driven assessments, and consistent evaluation.

How the Hiring Workflow Runs

AWS built the product for scaled hiring in industries such as retail, logistics, and hospitality, where, in the company’s description, recruiting teams fill hundreds of roles on tight timelines while stitching together applicant tracking systems, scheduling tools, and spreadsheets. Recruiters configure evaluation criteria, assessments, and interview questions based on each job’s requirements, and AI agents then conduct the interviews and assessments. AWS said the agents can process thousands of candidates across a hiring pipeline.

For candidates, an interview can be completed at any hour from any device, without a scheduling step. Recruiters return to a dashboard of scored candidates with complete transcripts and the evaluation reasoning behind every score, ready for review and a final decision. AWS said the product is informed by decades of Amazon’s hiring science, and the company described its intended effect as moving recruiters out of workflow management so organizations can fill roles before lost revenue and rising costs set in.

Evaluations, Interviews, and Assessments

An evaluation is the unit the product uses to measure a candidate for a job, and it can include an AI-led interview, assessments, or both, according to the Amazon Connect Talent Administrator Guide. A recruiter builds one through a step-by-step builder: after the recruiter selects or describes a job, the product generates a draft organized into an overview covering the job, its key competencies, and the evaluation language; a set of digital tests selected for those competencies; and an AI-led interview of behavioral questions grouped by competency. The recruiter reviews and adjusts the draft, then publishes it, and a published evaluation can be sent to candidates.

During an AI-led interview, the system asks behavioral questions generated for each competency the job requires, with dynamic follow-up questions that explore a candidate’s answers. Every question is tagged with the competency it measures and carries a short rationale, and questions can be generated in multiple languages.

Assessment results are reported on a four-level scale: Low, Moderate, High, and Very High. The Work approach assessment measures how a candidate approaches their work, and the Problem solving assessment measures how a candidate works through problems; for each, recruiters see a per-competency breakdown and an overall fit summary. When a candidate does not provide enough signal on a competency, the product excludes that competency from results rather than guessing.

Administrative features include applicant tracking system integration to synchronize jobs and candidates and return evaluation results, SAML-based single sign-on through an existing identity provider, role-based security profiles, knowledge bases that hold company values for personalizing evaluations, candidate-facing branding with an organization’s name, logo, and color, and data governance controls covering encryption, data residency, and retention.

Evaluation Standards, Security, and Availability

AWS said the AI in Connect Talent measures only job-related competencies, including problem-solving, logic, listening, and role-specific capabilities, and that candidate data is anonymized during AI evaluation. The company said the assessments are designed around competencies associated with success in the role rather than subjective impressions such as confidence level or speaking style. Each competency is scored against a rubric that defines strong and weak responses, and every score is tied to specific evidence from the interview so a result can be traced back to what the candidate said. Candidates are told before the evaluation what data is collected, how it is used, and what is not collected.

Recruiters retain final decision authority over every hire, with scored summaries, competency breakdowns, full transcripts, comparative analytics, and score reasoning in a single view. As an example of the workflow, AWS described a retail organization ramping up 500 warehouse roles for peak season, with AI agents interviewing candidates around the clock while recruiters review structured, evidence-based evaluations each morning.

On security, AWS described the product as enterprise-grade, built on AWS infrastructure with access that is controlled, auditable, and configurable. Integrity monitoring uses text-based analysis to flag unnatural cadence, filler words, pauses, and response latency. Human review is mandatory for every flag, no candidate is disqualified by automated signals alone, and recruiters control how flagged interviews are handled. Every candidate interaction is logged with a complete audit trail and job-related evaluation reasoning.

Amazon Connect Talent is available in the US East (N. Virginia) and US West (Oregon) AWS Regions, and customer data resides in the Region where the instance is created. The recruiter experience supports English, German, Spanish, French, Italian, Japanese, Korean, Portuguese (Brazil), and Simplified and Traditional Chinese; the candidate experience supports a separate set of locales covering US, Australian, and UK English, Portuguese (Brazil), French, Italian, German, and US Spanish. The product supports screen readers and keyboard navigation across both experiences, but the AI-led interview is not optimized for screen readers; candidates who use incompatible assistive technology can opt out of the interview, complete the rest of the evaluation, and keep their application Active in the pipeline.