BEGIN:VCALENDAR
VERSION:2.0
PRODID:Data::ICal 0.24
BEGIN:VEVENT
DESCRIPTION:   'Title: Overcast: Video OSINT Agent. Point It at 100 Videos\
 , Ask\n   Anything\n   Tags: Intro/Beginner | AI | DEF CON Demo Labs | Off
 ense/Red Team |\n   Threat Intel/Hunting | DEF CON Demo Labs\n   When: Sat
 urday\, Aug 8\, 16:00 - 16:45 PDT\n   Where: LVCCW Level 1 Hall 3 902 (Dem
 o Labs Track 6) - [1]Map\n\n   Description:\n\n   Conference talks\, earni
 ngs calls\, product demos\, training videos.\n   Organizations put hours o
 f footage online every week\, full of things\n   they didn't mean to share
 : hostnames in terminal windows\, org charts\n   on slides\, infrastructur
 e details dropped during Q&A. Traditional\n   OSINT can't touch video at s
 cale\, and manual review falls apart past a\n   handful of recordings.\n\n
    Overcast is a CLI agent and skill pack for video OSINT that drops into\
 n   any agentic harness\, such as Claude Code\, Codex\, or Tinycloud\, giv
 ing\n   it senses plus recon and targeting reach\, organized around an\n  
  investigation case. Point it at 10 videos or 1\,000 and it turns\n   foot
 age into cited evidence: speech\, video understanding\, on-screen\n   text
  and objects\, faces\, and named entities\, all accumulating in\n   persis
 tent case memory.\n\n   Discovery runs on the same case: scan and monitor 
 sweep sources and\n   surface reviewable findings. Ask across the whole co
 rpus and get\n   answers cited to the exact record and timestamp\, backed 
 by tiered\n   retrieval. Match a photo against thousands of clips to find 
 a person.\n   Each subcommand is modular and pluggable\, so analysts can d
 rop one\n   into other agent flows\, author custom skills\, or feed Overca
 st's media\n   analysis into existing recon and security tooling to comple
 te the\n   mission.\n\n   SpeakerBio:  Kevin "kdrwins" Dela Rosa\n\n   Kev
 in Dela Rosa is a multimodal AI researcher and engineer with 17+\n   years
  in computer vision\, NLU\, and large-scale retrieval. He led\n   engineer
 ing teams at Snapchat building billion-scale visual search and\n   generat
 ive AI products\, worked on large-scale ML at Amazon\, and\n   interned at
  NIST and SPAWAR on NLP and information retrieval for\n   government appli
 cations. He's published at ACM WWW\, ACM CAIS\, IEEE\n   ICCV\, KDD\, CVPR
 \, NeurIPS\, AAAI\, and ISMIR\, and has spoken at AWS\n   re:Invent\, Kube
 Con\, and CascadiaJS. He's currently CTO of Cloudglue\,\n   where he build
 s video understanding infrastructure.\n\n   At DEF CON 33 he presented "Au
 tonomous Video Hunter" at Recon Village\,\n   demoing an AI agent that inv
 estigated video corpora using face\n   recognition\, logo detection\, and 
 content analysis to produce\n   structured OSINT reports. Overcast is the 
 evolution of that work: a\n   CLI agent and skill pack that gives any agen
 tic harness senses plus\n   recon and targeting reach\, organized around a
 n investigation case with\n   persistent memory\, cited evidence\, and mod
 ular subcommands that plug\n   into other recon and security workflows.\n\
 n   Links:\n       GitHub - [2]https://github.com/kdr/overcast\n   '\n\n  
  1. #LVCCW_Level1_Hall3\n   2. https://github.com/kdr/overcast\n\n\n
DTEND:20260808T234500Z
DTSTART:20260808T230000Z
LOCATION:Demo Labs - LVCCW Level 1 Hall 3 902 (Demo Labs Track 6)
SUMMARY:Overcast: Video OSINT Agent. Point It at 100 Videos\, Ask Anything
END:VEVENT
END:VCALENDAR