Skip to main content

Performance and Reliability

Service level agreements, performance metrics, and reliability commitments for the botsKYC platform.

Service Level Agreement (SLA)​

Uptime Guarantee​

We commit to high availability for all services.

SLA Commitment: 99.9% Uptime

  • Monthly Uptime: 99.9% (4 minutes downtime/month)
  • Measurement Period: Calendar month
  • Exclusions: Scheduled maintenance (announced 7 days prior)

Uptime Calculation:

Uptime % = (Total Minutes - Downtime Minutes) / Total Minutes × 00

Response Time​

Fast processing for real-time and batch operations.

Synchronous API (Real-Time)​

Target: < seconds (95th percentile)

EndpointTargetMeasured
Identity Verification< s.8s
Address Verification< s.9s
Income Verification< .5s.s
Liveness Session< s0.5s
Liveness Results< .5s.s

Asynchronous API (Batch)​

Target: < 0 seconds (95th percentile)

OperationTargetMeasured
Batch Verification< 0s8s
Complex Documents< 45ss
Multi-Page PDFs< 60s4s

Availability by Region​

RegionSLAActual (Last 0 Days)
Botswana99.9%99.95%
Global99.9%99.9%

Performance Metrics​

Throughput​

Maximum requests the system can handle.

Capacity:

  • Per Account: ,000 requests/minute
  • Burst Capacity: ,000 requests/minute (0 seconds)
  • Enterprise: Custom limits available

Actual Performance:

  • Average: 850 req/min
  • Peak: ,800 req/min
  • Sustained: ,00 req/min

Latency​

End-to-end response times by percentile.

API Response Times (ms)​

PercentileIdentityAddressIncomeLiveness
50th (Median),00,400,500800
75th,500,700,900,000
95th,800,900,00,00
99th,400,600,800,800

Geographic Latency​

RegionConnectionAPI ProcessingTotal
Within Botswana0-50ms,00ms,0-,50ms
Southern Africa50-00ms,00ms,50-,00ms
Europe50-00ms,00ms,50-,400ms
North America00-50ms,00ms,400-,450ms
Asia50-50ms,00ms,450-,550ms

Success Rates​

Percentage of successful operations.

Target: > 99.5% Success Rate

ServiceTargetActual
Document Verification99.5%99.7%
Liveness Detection99.5%99.8%
Data Extraction99.5%99.6%
Face Matching99.5%99.7%

Error Categories:

  • Client Errors (4xx): 0.%
  • Server Errors (5xx): 0.%
  • Timeouts: < 0.%

Reliability Features​

High Availability​

Multiple redundancy layers.

Infrastructure​

  • Multi-Zone Deployment - Across + availability zones
  • Load Balancing - Traffic distributed automatically
  • Auto-Scaling - Dynamic capacity adjustment
  • Health Checks - Every 0 seconds

Failover​

  • Automatic Failover - < 0 seconds
  • Manual Override - Available 4/7
  • Rollback Capability - Instant version rollback
  • Circuit Breakers - Prevent cascade failures

Data Durability​

Protection against data loss.

Storage Durability: 99.999999999% ( nines)

  • Replication - + copies across zones
  • Automated Backups - Every 6 hours
  • Point-in-Time Recovery - 5-day retention
  • Geo-Replication - Cross-region backups

Backup Schedule:

  • Continuous - Transaction logs
  • Hourly - Incremental backups
  • Daily - Full backups
  • Weekly - Long-term archives

Disaster Recovery​

Recovery Objectives​

  • RTO (Recovery Time Objective) - < 5 minutes
  • RPO (Recovery Point Objective) - < minute
  • Data Loss - Near-zero

DR Capabilities​

  • Multi-Region Backup - Geographically distributed
  • Automated Recovery - No manual intervention
  • Failover Testing - Quarterly DR drills
  • Runbooks - Documented procedures

Monitoring and Alerting​

Real-Time Monitoring​

4/7 system monitoring.

Monitored Metrics:

  • API response times
  • Error rates
  • Throughput
  • System resources
  • Queue depths
  • Cache hit rates

Alert Thresholds:

  • Response time > s
  • Error rate > %
  • Uptime < 99.9%
  • Queue depth > 000

Status Page​

Public system status updates.

Access: status.botskyc.com

Information:

  • Current system status
  • Incident history
  • Scheduled maintenance
  • Performance metrics

Subscription Options:

  • Email notifications
  • SMS alerts
  • Webhook updates
  • RSS feed

Incident Response​

Response Times​

SeverityResponse TimeResolution Target
Critical< 5 minutes< hour
High< hour< 4 hours
Medium< 4 hours< 4 hours
Low< 4 hours< 7 hours

Communication​

  • Status Page - Real-time updates
  • Email Notifications - Subscribed users
  • API Status Endpoint - Automated checks
  • Support Channel - Direct contact

Performance Optimization​

Caching​

Multi-layer caching strategy.

Cache Layers: . CDN Cache - Static content (95% hit rate) . API Gateway Cache - Response caching (60% hit rate) . Application Cache - Business logic (80% hit rate) 4. Database Cache - Query results (75% hit rate)

Benefits:

  • Reduced latency
  • Lower server load
  • Cost efficiency
  • Better scalability

Rate Limiting​

Fair usage and protection.

Limits:

  • Standard: ,000 req/min
  • Pro: ,500 req/min
  • Enterprise: Custom limits

Headers:

X-RateLimit-Limit: 000
X-RateLimit-Remaining: 950
X-RateLimit-Reset: 699894

Exceeded Response:

{
  "error": "rate_limit_exceeded",
  "message": "Too many requests",
  "retryAfter": 60
}

Content Delivery​

Global CDN for fast asset delivery.

CDN Features:

  • 00+ edge locations
  • Automatic compression
  • HTTP/ support
  • DDoS protection

Performance:

  • 90% of users < 50ms
  • 99% of users < 00ms
  • Average: 45ms

Capacity Planning​

Resource Allocation​

Dynamic resource management.

Auto-Scaling Triggers:

  • CPU > 70%
  • Memory > 80%
  • Queue depth > 500
  • Response time > s

Scaling Actions:

  • Add instances (< minutes)
  • Increase memory allocation
  • Optimize database connections
  • Enable read replicas

Load Testing​

Regular performance validation.

Test Frequency: Monthly

Test Scenarios:

  • Normal load (,000 req/min)
  • Peak load (5,000 req/min)
  • Sustained load (4 hours)
  • Spike testing (sudden 0x increase)

Success Criteria:

  • Response time < SLA targets
  • Error rate < 0.5%
  • No system crashes
  • Graceful degradation

SLA Credits​

Credit Policy​

Compensation for service disruptions.

Monthly Uptime SLA:

Uptime %Service Credit
< 99.9%0%
< 99.0%5%
< 95.0%50%
< 90.0%00%

How to Claim: . Submit ticket within 0 days . Provide verification details . Credits applied next billing cycle

Exclusions:

  • Scheduled maintenance
  • Client's network issues
  • Third-party service failures
  • Force majeure events

Best Practices​

Optimize Performance​

Tips for best results.

Use appropriate endpoints - Identity vs. multipurpose
Batch requests - Group multiple documents
Implement caching - Cache verification results
Async for bulk - Use webhooks for batches
Retry logic - Exponential backoff
Monitor usage - Track rate limits

Avoid polling - Use webhooks instead
Don't retry immediately - Wait between retries
Large files - Compress before upload

Handle Errors​

Robust error handling.

async function verifyDocument(file) {
  const maxRetries = ;
  let attempt = 0;

  while (attempt < maxRetries) {
    try {
      return await api.verify(file);
    } catch (error) {
      if (error.status === 49) {
        // Rate limited - wait and retry
        await sleep(error.retryAfter * 000);
      } else if (error.status >= 500) {
        // Server error - retry with backoff
        await sleep(Math.pow(, attempt) * 000);
      } else {
        // Client error - don't retry
        throw error;
      }
      attempt++;
    }
  }
  throw new Error('Max retries exceeded');
}

Support and Resources​

Status and Monitoring​

  • Status Page: status.botskyc.com
  • API Health: GET /api/health
  • Ping Endpoint: GET /api/ping

Documentation​

Contact​

  • Email: support@botskyc.com
  • Emergency: Available 4/7 for critical issues
  • Response Time: < hour for urgent matters

Next Steps​