AlektroAI

AI Security API

Every security capability. One API.

Threat detection, anomaly scoring, content moderation, vulnerability scanning, malware classification, and authentication risk scoring — all programmatically accessible with sub-100ms latency and full explainability.

Get API access

API reference

Endpoints

POST/v1/threat/analyze
POST/v1/anomaly/score
POST/v1/content/moderate
POST/v1/vulnerability/scan
POST/v1/malware/classify
POST/v1/auth/risk
GET/v1/threat/intel/{ioc}
GET/v1/audit/logs
POST /v1/threat/analyze

Threat Detection

Real-time analysis of logs, network traffic, and payloads for malicious patterns. Aligned to MITRE ATT&CK with confidence scores and SHAP-based explanations on every result.

  • Inline inspection under 100ms for real-time use cases
  • MITRE ATT&CK-aligned detection coverage
  • SHAP explainability — flagged because of X pattern
  • Custom YAML detection rules for your environment
  • SIEM integration via webhook or Kafka
POST /v1/anomaly/score

Anomaly Detection

Behavioral baselines and deviation scoring. Score any behavioral event — API calls, login patterns, data access — against established baselines to surface threats before they escalate.

  • Automatic baseline learning from historical traffic
  • Deviation scoring with configurable sensitivity
  • Time-series storage via InfluxDB / TimescaleDB
  • Vector similarity matching for novel threat patterns
  • Burst handling for DDoS and incident surges
POST /v1/content/moderate

Content Moderation

Classify harmful, abusive, or policy-violating content at scale. Configurable confidence thresholds and human-in-the-loop review flows reduce false positives and alert fatigue.

  • 40+ content violation categories out of the box
  • Configurable confidence thresholds per policy
  • Human-in-the-loop review flow for borderline cases
  • GDPR / CCPA-compliant — data minimization enforced
  • Bidirectional input and output inspection
POST /v1/vulnerability/scan

Vulnerability Scanning

Analyze code and configuration files for known CVEs and misconfigurations. CI/CD-native with GitHub Actions, GitLab CI, and Jenkins — catch vulnerabilities before they reach production.

  • CVE database updated continuously via threat intel feeds
  • Misconfiguration detection for cloud and container configs
  • CI/CD pipeline integration — blocks on critical findings
  • OWASP LLM Top 10 coverage for AI-facing APIs
  • On-demand and scheduled scan modes
POST /v1/malware/classify

Malware Classification

Static and dynamic analysis of files and binaries. Multi-model ensemble powered by TorchServe / Triton delivers high-confidence verdicts with reasoning you can act on immediately.

  • Static analysis — PE, ELF, Mach-O, scripts, documents
  • Dynamic behavioral sandboxing for unknown samples
  • Multi-model ensemble for high-confidence verdicts
  • Async webhook mode for large file batches
  • VirusTotal and MISP threat intel enrichment
POST /v1/auth/risk

Auth Risk Scoring

Evaluate login attempts for fraud signals in real time. Risk scores power step-up authentication decisions without adding friction to legitimate users.

  • Device, location, and behavioral signal fusion
  • Real-time scoring under 50ms for inline auth gates
  • OAuth 2.0 integration — drop into any auth flow
  • Adaptive thresholds via continuous model retraining
  • Explainable output — know exactly why a login was flagged

Architecture

Key design decisions

Every default is chosen for production security workloads. Overrides available for all tiers.

DecisionOptionsOur default
ProtocolREST vs gRPCgRPC for low-latency paths; REST for broad compatibility
AuthenticationAPI keys vs OAuth 2.0OAuth 2.0 for enterprise; API keys for quick integration
Response modeSync vs async webhookSync for real-time inline checks; async for deep scans
Model hostingSelf-hosted vs cloudSelf-hosted for sensitive data; cloud for ease of scale
ExplainabilitySHAP, LIME, attention mapsSHAP for tabular signals; attention maps for NLP models

Start building with the AI Security API

Join hundreds of engineering and security teams who rely on AlektroAI for real-time threat detection and compliance.

Get API access