81%
Lower screening costs
Loading Decision Agent...
Solution
Decision Agent is recruitment assessment software for agencies, internal teams and hiring managers. Start with a role brief, choose the assessment stages and invite candidates to complete them. Review their answers, work and conversation evidence before deciding what happens next.
Automate the work of creating and running early-stage assessments. Spend recruiter time inspecting the evidence and deciding what it means.
1. Create the assessment
Provide a job description or brief. Decision Agent drafts the assessment and behaviours to look for; refine every element with the operator before using it.
2. Invite candidates
Send email invitations to a mobile-accessible assessment. Candidates answer the questions they need to answer, work through a text scenario or speak with an AI roleplayer, depending on your stages.
3. Inspect the evidence
Review questionnaire answers and behaviour findings linked to the scenario response or conversation transcript. Conversation reports include playable clips of the supporting exchange.
4. Review and decide
Challenge findings, apply your criteria and decide what happens next. Pinpoint can receive progression or rejection decisions and send status updates back to Decision Agent.
Use a job description or brief to draft the assessment, then refine the situations, objectives and behaviours with the operator. The criteria are agreed before candidates respond.
Assessment design
Operations lead
Illustrative configuration
Role brief
Added
Behaviours
Defined
Stages
Reviewed
Behaviour to observe
Clarifies constraints before committing to an action.
Assessment opportunity
A work situation with incomplete information and competing requests.
Decision Agent
Clarify the role requirements
Review the proposed behaviours
Refine the assessment stages
These figures were reported from a launch client’s comparison with its previous screening services, involving around 3,000 applicants. They describe that deployment and do not establish assessment validity or predict results for every customer.
81%
Lower screening costs
3x
Time-to-hire improvement
96%
Candidate satisfaction
Evidence first
Inspect what the candidate did, the predefined behaviour being assessed and the explanation for the finding. Use that evidence in human review.
See it in Action
For Scenarios and Conversations, AI checks whether a predefined behaviour is fully or partially demonstrated in a particular response or transcript. The report shows the finding, its explanation and the evidence supporting it. Recruiters can inspect and correct individual findings.
Where a score is shown, fixed rules combine the behaviour findings using recruiter-defined weights. That score is guidance: it does not establish overall suitability or predict future job performance.
How to read the evidenceAs assessment responses and later outcomes become available, the aim is to study which behaviours relate to success in the role, including evidence from performance reviews. That future work could inform revised behaviours and weights. It is not a current prediction capability.
Validating the automation of an individual questionnaire question is a separate task. The team uses human-labelled answers and tailored machine-learning methods to meet the validation target agreed with the recruiter.
Explore the existing assessment examples and sample reports in the brochure, then read the full product guides in the Library.
Legal, Academia and AI Aptitude are available configurations of the shared assessment modules.
Candidates can email support@decision-agent.ai for technical help or to raise a difficulty completing an assessment. The hiring recruiter decides any exception. Support does not change the scenario’s work or assessment criteria.