Getting Started

  • Introduction
  • Concepts & Glossary
  • Quickstart
  • Authentication

API Reference

  • Endpoints
  • Error Handling
  • Versioning

Guides

  • Code Examples
  • FHIR R4 Integration
  • Integration Patterns
  • Going to Production

Medical Content

  • Knowledge Base

Tools

  • Interactive Sandbox
  • API Status

Resources

  • FAQ
  • Changelog
  • Support
DevelopersGuidesGoing to Production

Loading documentation…

Integration PatternsKnowledge Base
PearMedica Logo

The clinical intelligence layer built for African healthcare. Empowering providers with AI-driven triage and decision support.

Product

  • Symptom Assessment
  • Triage Engine
  • Decision Support
  • Pricing

Developers

  • Documentation
  • API Reference
  • Quickstart Guide
  • Code Examples

Company

  • About Us
  • Contact
  • Blog
  • Careers

Legal

  • Privacy Policy
  • Terms of Service
  • Security & Compliance
  • NDPA Compliance
NDPA 2023
AES-256 Encryption
Audit Logging

© 2026 PearMedica. All rights reserved.

Status

Going to Production

A checklist for launching your PearMedica integration. Use this guide to ensure your deployment is secure, reliable, and clinically safe before going live.

API Keys & Security

!

Switch from sk_test_ to sk_live_ keys

CRITICALCritical
!

Store API keys in environment variables — never in source code

CRITICALCritical

Use separate API keys for staging and production

Enable key rotation schedule (quarterly recommended)

!

Restrict API calls to server-side only — never expose keys in client-side code

CRITICALCritical

Error Handling

!

Implement retry logic with exponential backoff for 429/500/503 errors

CRITICALCritical

Add circuit breaker pattern (stop calling after 5 consecutive failures)

!

Build a fallback UI for when the API is unavailable

CRITICALCritical

Log all API errors with request context for debugging

Handle QUOTA_EXCEEDED (403) gracefully — show upgrade prompt

Monitoring & Observability

!

Poll GET /v1/health every 30 seconds to monitor API availability

CRITICALCritical

Track assessment latency (p50, p95, p99) in your monitoring stack

Set up alerts for error rate spikes (>1% of requests failing)

Monitor rate limit headers (X-RateLimit-Remaining) and alert at 80% utilisation

Review GET /v1/usage regularly to track quota consumption

Clinical Safety

!

Display the medical disclaimer on every assessment shown to patients

CRITICALCritical
!

Handle "emergency" triage responses with immediate escalation to human staff

CRITICALCritical
!

Never present AI assessments as definitive diagnoses

CRITICALCritical

Log all emergency-level assessments for clinical audit

!

Test red flag scenarios (chest pain + shortness of breath) to verify emergency detection

CRITICALCritical

Switching from Test to Live

1

Verify in Sandbox

Ensure all integration tests pass using your sk_test_ key. Verify all edge cases: empty evidence, emergency patterns, rate limit handling.

2

Generate Production Key

Go to Dashboard → API Keys → Create New Key → Select "Production". Save the key securely — it is shown only once.

3

Update Environment Variables

Set PEARMEDICA_API_KEY=sk_live_your_key in your production environment. Never commit this to version control.

4

Deploy & Monitor

Deploy your application. Monitor the first 100 assessments for error rates, latency, and triage distribution. Check the Dashboard for real-time usage.

NDPA 2023 Compliance Note

If your application processes real patient data in Nigeria, ensure your setup complies with the Nigeria Data Protection Act 2023. PearMedica's MVP uses synthetic data on Supabase. Post-funding, production data will be migrated to Nigerian-hosted infrastructure. Contact us for guidance on compliance requirements.

Related Resources

Error Handling — retry strategies and graceful degradationAuthentication — API key format and rate limits