Score Documentation
AI-driven evaluation portal for scoring tenders, recruitment, grants, and more
What is Score?
Score is an AI-powered evaluation platform that helps you objectively assess submissions, candidates, proposals, and applications. Define your criteria, upload documents, and let AI provide consistent, evidence-based scoring recommendations.
Key Features
- Multi-Candidate Scoring: Compare multiple candidates, proposals, or submissions simultaneously against the same criteria
- Criteria Matrices: Create weighted, points-based, or pass/fail evaluation frameworks with custom thresholds
- Template Library: Pre-built domain models for recruitment, tenders, grants, and vendor selection
- Document Analysis: Upload PDFs, Word docs, or paste text for AI-driven evaluation
- Position Description Upload: For recruitment evaluations, upload PD documents to provide comprehensive role context
- Comparison View: View all candidates ranked by score with detailed criteria breakdowns
- Shortlisting Tools: Automatically select top N candidates or manually curate your shortlist
- Manual Override: Override AI decisions with audit trail for compliance
- Evidence Extraction: AI provides direct quotes and specific evidence from documents
- Team Collaboration: Share evaluations across your organization
How It Works
Single Submission Evaluation
- Create an Evaluation: Start from a template or build custom criteria with weightings
- Upload Document: Add a single document (PDF, Word doc, or paste text) to be evaluated
- Run AI Analysis: AI extracts key information and scores against your criteria
- Review Results: Get detailed scores, justifications, and evidence for each criterion
- Manual Override: Override any AI decisions if needed with documented reasoning
Multi-Candidate Evaluation (Recruitment)
- Create Evaluation: Select the Recruitment template with 6 pre-configured criteria
- Upload Position Description (Optional): Provide role context for more accurate AI assessment
- Configure Criteria: Customize scoring criteria, weights, and AI guidance
- Submit Candidates: Add multiple candidates with their individual documents (resumes, cover letters)
- Run AI Analysis: AI scores each candidate against all criteria simultaneously
- View Comparison: See all candidates ranked by score with criteria breakdown in table or card view
- Shortlist: Auto-select top N candidates or manually curate your shortlist
- Export: Copy shortlist or export detailed comparison report
Use Cases
- Recruitment Shortlisting: Score multiple candidate applications simultaneously to identify top talent for interviews. Upload the position description and evaluate all resumes against standardized criteria including qualifications, experience, references, and police checks.
- Tender Evaluation: Score vendor proposals against technical, pricing, and delivery criteria. Compare multiple submissions to identify the best value provider.
- Grant Assessment: Evaluate multiple grant applications for merit, feasibility, budget alignment, and impact. Rank applicants to make funding decisions.
- Vendor Selection: Compare multiple service providers on quality, cost, support, and reliability to select the optimal partner.
- RFP Responses: Systematically evaluate multiple responses to requests for proposal, ensuring consistent scoring across all submissions.
- Compliance Checking: Use pass/fail criteria to verify regulatory requirements across multiple submissions (e.g., insurance certificates, certifications).
Scoring Models (Templates)
Score includes pre-configured domain models with industry-standard criteria and AI guidance:
- Recruitment Assessment (Pass/Fail): 6 criteria designed for candidate shortlisting
- IT Degree - Verification of IT or related field degree
- Police Check - Evidence of background check documentation
- Adobe Creative Cloud Skills - Demonstrated proficiency
- Right to Work - Eligibility verification
- Valid Driver's License - License confirmation
- Salary Expectations - Budget alignment check
- Tender Evaluation (Weighted): Technical capability (30%), Pricing (25%), Delivery (20%), References (15%), Innovation (10%)
- Grant Assessment (Weighted): Project merit (30%), Feasibility (25%), Budget (20%), Impact (15%), Sustainability (10%)
- Vendor Selection (Points): Service quality, Pricing, Support, Reliability, Integration
All templates can be customized to fit your specific needs. Modify criteria, adjust weightings, or add domain-specific AI guidance.
Getting Started
Quick Start - Recruitment Example
- Navigate to Score from your Launchpad
- Click "+ New Evaluation" from the Dashboard
- Step 1 - Basic Information:
- Give your evaluation a descriptive title (e.g., "Web Developer - January 2025")
- Select "Recruitment" as the evaluation type
- Optionally upload the Position Description PDF for enhanced AI context
- Fill in role details (job title, department, required qualifications, salary range)
- Step 2 - Scoring Criteria:
- Select the "Recruitment Assessment" scoring model
- Review the 6 suggested pass/fail criteria
- Customize AI guidance for each criterion if needed
- Mark mandatory criteria (e.g., Right to Work, Police Check)
- Step 3 - Submit Candidates:
- Click "+ Add Candidate" to add each applicant
- Enter candidate name, email, and reference number
- Upload their resume/CV and cover letter, or paste text
- Repeat for all candidates (you can add/edit/remove anytime)
- Start Scoring: Click "Save & Start Scoring" to begin AI analysis
- View Results: Monitor progress and view individual candidate scores as they complete
- Compare & Shortlist: Click "View Comparison" to see all candidates ranked together
- Make Decisions: Use auto-shortlist to select top N candidates or manually curate your list
Best Practices
- Define Clear Criteria: Use specific, measurable criteria before starting evaluations. Vague criteria lead to inconsistent AI scoring.
- Customize AI Guidance: Each criterion can have custom AI instructions. Be explicit about what constitutes pass/fail or high/low scores.
- Use Position Descriptions: For recruitment, uploading the PD provides crucial role context that improves AI accuracy significantly.
- Multi-Candidate Mode: When comparing multiple submissions (recruitment, grants, tenders), use multi-candidate mode for consistent scoring across all applicants.
- Review AI Evidence: Always review the evidence quotes and reasoning provided by AI before making final decisions. AI highlights what it found in the documents.
- Manual Override When Needed: If you have information not in the documents (e.g., sighted police check in person), use manual override with clear reasoning for audit trail.
- Save Scoring Models: Create reusable scoring models for recurring evaluation types (e.g., quarterly recruitment for same role).
- Consistent Weightings: Use consistent weightings across similar evaluations to ensure fair comparisons.
- Document Format: PDFs and Word docs (.docx) work best. Ensure text is selectable (not scanned images) for optimal AI extraction.
Tips & Troubleshooting
- Police Check False Positives: The AI has been specifically trained to distinguish between "references available" and actual police check documentation. Only explicit mention of background checks counts as valid.
- Scoring Taking Too Long: Multi-candidate evaluations process each candidate sequentially. Scoring time = (candidates × criteria × ~5 seconds). A 5-candidate, 6-criteria evaluation takes approximately 2-3 minutes.
- Viewing Progress: The evaluation page shows real-time progress. You can navigate away and return - scoring continues in the background.
- Comparison View Not Showing: The "View Comparison" button only appears for multi-candidate evaluations that have status "completed" or "failed".
- Export Options: Currently you can copy the shortlist to clipboard. CSV and PDF export features are coming soon.