Insurance × Agentic AI × Cybersecurity × Physical AI
Joe Murillo / Insurance • Risk Engineering • Intelligent Systems
My work spans insurance business development and loss control. I bring that field perspective to projects in agentic AI, cybersecurity, and physical AI.
01 / Intelligent Risk Engineering
Four connected disciplines. One evidence-led approach.
Risk / Engineering focus
Field observation, source validation, and accountable loss-control reasoning.
02 / Intelligent Risk Engineering
Inspect scope, source material, limitations, and the next verifiable step.
Governed systems / Editorial imageryGoverned agent workflows with mission lifecycle checks, per-call authorization, and local governance history.
Public documentation and selected source reviewed; runtime behavior has not been independently verified in this website review.
Commercial growth intelligence positioning for construction, building materials, manufacturing, and distribution.
Open dossier →A proposed governance and oversight layer connecting authorization, review, and accountable decisions.
Open dossier →A network-forensics learning workflow extracting DNS, TLS SNI, and protocol statistics for analyst review.
Open dossier →A static Markdown security-briefing scaffold; export ingestion and live integrations remain future work.
Open dossier →03 / Intelligent Risk Engineering
Curated public repositories. Scope comes from reviewed documentation; GitHub remains the source of current activity.
GitHub metadata snapshot observed 2026-10-16; manually reviewed before inclusion. Repository push dates do not establish shipped features. Open GitHub for current activity.
04 / Intelligent Risk Engineering
An interactive conceptual model, not a live risk score. Select a discipline to inspect its relationships.
Insurance / Conceptual relationships
Translate an observation into questions about exposure, uncertainty, controls, and accountable decisions.
Conceptual relationships only. No scoring engine, live integration, or causal model is claimed.
05 / Intelligent Risk Engineering
Public milestones with explicit provenance. These entries describe this catalog’s evidence review, not private operational events.
Downgraded the public description from export-first MVP to static report scaffold after inspecting the script. Recorded the README/code mismatch and future ingestion work.
Read the evidence boundary →Recorded public lifecycle, authorization, and persistence contracts with their limitations. This is a documentation review, not an execution test.
Read the evidence boundary →Separated the existing landing-page prototype from future intelligence-engine ambitions.
Read the evidence boundary →Cataloged public forensic and security-briefing workflows; retained explicit analyst-review boundaries.
Read the evidence boundary →06 / Intelligent Risk Engineering
From network evidence to physical observations. Experiments are separated from demonstrated capability.
Workstation and network experiments; no public inventory or validated benchmark is claimed.
PCAP triage documents evidence extraction and analyst reporting.
ARIS documents governed agent workflows with local authorization boundaries.
Drones, sensors, and computer vision for inspection evidence; no operational flight or model is demonstrated.
Interactive demonstration / Synthetic data
Work through a fictional field observation. The review gate must be acknowledged before recording a synthetic outcome.
Fictional observation: dust accumulation near a woodworking station. Record time, location, and equipment context.
07 / Intelligent Risk Engineering
Engineering perspectives on evidence, authority, and real-world risk.
A decision-support system becomes useful when a reviewer can reconstruct why it reached a conclusion.
Read field note →Research framework / 2026-10-16Connect field observations to accountable recommendations before adding automation.
Read field note →08 / Intelligent Risk Engineering
Explore the reviewed public knowledge base with source citations.
Public knowledge / Local retrieval
Answers are retrieved from the approved catalog. This preview uses deterministic retrieval, not a generative model. Your question stays in this browser; nothing is sent to an AI provider.
Answers and approved citations appear here. No account, upload, or operational access is required.
