Sentinel AI: Concurrent Threat Intelligence & Explainable Triage
In Security Operations Centers (SOCs), analysts face a critical bottleneck: alert fatigue. Every incoming security alert requires manually correlating domain registrations, VirusTotal reputation scores, autonomous system numbers (ASNs), and historical DNS records.
I designed Sentinel AI to compress this multi-hour investigative workflow into sub-second concurrent telemetry analysis with Explainable AI triage.
Concurrent OSINT Gathering
Sequential API calls to external security databases introduce unacceptable latency. Sentinel AI utilizes Python's asyncio and httpx to trigger parallel telemetry workers:
- VirusTotal v3 Endpoint: Querying multi-engine detection ratios and behavioral sandbox reports.
- WHOIS & RDAP: Extracting registrar creation dates, registrant privacy indicators, and domain age.
- DNS Telemetry: Resolving active A, AAAA, MX, and TXT records to detect fast-flux infrastructure.
async def investigate_indicator(ioc: str, ioc_type: IOCType) -> TelemetryReport:
async with httpx.AsyncClient(timeout=10.0) as client:
vt_task = fetch_virustotal_report(client, ioc)
whois_task = query_whois_record(client, ioc)
dns_task = resolve_dns_records(ioc)
vt_res, whois_res, dns_res = await asyncio.gather(
vt_task, whois_task, dns_task, return_exceptions=True
)
return compile_ioc_verdict(vt_res, whois_res, dns_res)
Explainable AI (XAI) Scoring
Rather than presenting analysts with an opaque "Risk Score: 87/100", Sentinel AI uses an explainable heuristic attribution engine. Each risk modifier is tied to a specific evidence anchor:
+35 PTS: Domain registered less than 7 days ago.+40 PTS: 5+ antivirus vendors flagged hash as malicious ransomware.+15 PTS: Mismatched SSL/TLS certificate subject alternative names (SAN).
By presenting the exact signals that contributed to the score, junior analysts can immediately validate findings without second-guessing black-box models.
Conclusion
Modern cyber defense requires combining rapid asynchronous data aggregation with transparent, explainable triage. Sentinel AI proves that automating tier-1 triage saves valuable minutes during active incident response.
Explore the source on GitHub.