Core features for fair, lawful, and explainable AI hiring.
TraceAI sits downstream of your AI tools. It doesn't make decisions — it logs, explains, monitors for bias, and generates the compliance documentation those tools lack.
Connect your AI stack in five lines
TraceAI publishes SDKs on PyPI and npm under gettraceai. Install, authenticate with your API key, and start logging decisions. No procurement cycle, no IT project.
- Python and Node.js SDKs — published, versioned, documented
- Single endpoint:
POST /v1/decisions - Batch support — up to 100 decisions per call
- Sandbox and production isolation with
sk_test_/sk_live_keys
from traceai import TraceAI
client = TraceAI(api_key="sk_live_...")
client.log_decision(
candidate_id="c_8291",
role="Senior Engineer",
ai_recommendation="shortlist",
confidence=0.87,
human_decision="shortlist",
model="gpt-4o",
factors=["experience", "skills"]
)
Real-time decision monitoring and bias detection
Every decision flows into a dashboard where your compliance team can track outcomes, spot patterns, and catch bias before a candidate or regulator does.
- Four-fifths rule analysis across gender, age, and ethnicity
- Per-role pass/fail rate breakdown with flag distribution
- Ethnic clustering detection and age confidence gap analysis
- Automated email alerts when thresholds are breached
Audit-ready compliance reports in one click
TraceAI generates 8-section PDF compliance reports aligned to UK GDPR Article 22, the ICO Statutory Code, and EU AI Act high-risk requirements. Built from your real decision data — not templates.
- Executive summary with compliance readiness score
- Regulatory coverage matrix across UK GDPR, EU AI Act, ICO Code
- Decision activity analysis with bias monitoring results
- Audit trail integrity verification with hash chain status
- Gaps and recommendations section with remediation steps
Log a Decision
Records a single AI-assisted hiring decision with candidate demographics, model metadata, and human oversight status.
-H "x-api-key: sk_live_..." \
-H "Content-Type: application/json" \
-d '{"candidate_id": "c_8291", ...}'
Complete API reference at /docs
Every endpoint, payload schema, and response format documented with working code examples in Python and Node.js. Webhook setup, SDK installation, authentication — everything your engineering team needs to integrate in a day.
- Decision ingestion, batch import, and retrieval endpoints
- Webhook configuration with HMAC-SHA256 signature verification
- SDK quickstart guides for Python and Node.js
- Authentication, API key management, and sandbox/production modes
Ready to make your AI hiring auditable?
Join our design partner programme and get full platform access free for three months.
Talk to us