A.E.G.I.S
Automated Extraction, Gathering, and Intelligence System — 2026, in progress
End-to-End Developer (system design, database, deployment) — self-initiated project
Most organizations only find out about a damaging public narrative after it's already gone viral — by then, every response looks reactive instead of thoughtful. A.E.G.I.S is a working OSINT and security intelligence platform I built end-to-end: FastAPI and Celery on the backend, Next.js on the frontend, PostgreSQL with pgvector for entity storage.
It runs sentiment analysis on Indonesian social text using IndoBERT, maps relationships between accounts and domains as an entity graph, and integrates natively with Wazuh SIEM for host-level security correlation — alongside a set of OSINT reconnaissance tools (Shodan, certificate transparency, ASN/BGP, WHOIS, dark web monitoring). Every data collection method stays within legal and platform boundaries — no paid APIs, no access bypassing.
Impact — Every module is live and running on real data — IndoBERT sentiment analysis on real Indonesian text, Wazuh SIEM correlating real security alerts. The real value isn't the alert itself, it's the lead time it gives a Relationship Officer or Relationship Manager to plan a proportionate response before a narrative goes public.
