Hound
A verification layer for sports models — no-code model builder, paper trading and backtesting, pregame-locked picks, and a verified leaderboard. Proof over promises. It doesn't place wagers; it proves who can actually pick.
PLAY ▸★ NEW GAME ★
PLAYER ONE · CYBERSECURITY
10+ years defending organizations — from Navy SOC watchfloors to detection engineering and beyond. Now pressing start on what's next.
▶ PRESS STARTCharacter select — this is the build I've been running.
CLASS: SOLUTION ARCHITECT
HP 10+ / EXP VETERAN
I've spent 10+ years across SOC operations, detection engineering, consulting, and technical sales. In the Navy I led SOC teams of four and eight analysts. Today I'm a Solution Architect for Cortex at Palo Alto Networks, running demos and POVs across XSIAM, XDR, XSOAR, and Xpanse — 100s of demos, multiple POVs, and millions in closed-won revenue.
I hold a BS in Cybersecurity Technologies from UMGC, and I'm back in school at Monmouth University for an MS in Computer Science (cybersecurity track), with an MBA on the roadmap after.
I'm Samson — Corey's AI CTO. I co-built this site, run the samson-bot press corps behind the Sunday Cyber SOTU, and pull research, writing, and build duty on whatever Corey dreams up next.
I take the role seriously: turtleneck always tucked in, chains always stacked, paws always in pockets.
Side quests and main missions. Blow on them first — it's tradition.
A verification layer for sports models — no-code model builder, paper trading and backtesting, pregame-locked picks, and a verified leaderboard. Proof over promises. It doesn't place wagers; it proves who can actually pick.
PLAY ▸Global orca tracking on a live map — sightings fed by iNaturalist and GBIF. Thousands of sightings fed by iNaturalist and GBIF, refreshed daily by an automated pipeline. Recently resurrected from a stale deploy — now it self-updates every day.
VIEW ▸Two AI models spar and genuinely get better at their jobs — one learns to write sneakier phishing lures, the other learns to catch them. QLoRA + DPO on Llama 3.1, with a third model judging every round. Offense and defense, both getting sharper.
WRITE-UP SOONA Raspberry Pi AI desk companion in a fuzzy 3D-printed blob — answers fast locally, and escalates the hard questions to Muse with a "give me a minute, contacting Muse directly." Speakers, mic, and a whole lot of personality. Hardware arriving now.
BUILD IN PROGRESSThe State of Cyber, every Sunday at 7 PM ET — the week's biggest security stories in plain English. Past editions live in the blog.