Lexius

AI Security Guards

Winter 2026ActiveB2BSecurityArtificial IntelligenceSaaSComputer VisionSecurityAISan Francisco, CA, USA

Company

https://www.lexius.ai

Lexius turns any retail security camera into an AI security guard that detects shoplifting, identifies repeated and organized crime, and automates case generation with video evidence. Lexius is the AI layer for existing security cameras. No rip and replace. No new hardware. No expensive upgrades. Think of it as a security guard that never misses a moment.

Verdict

High Signal

Market Opportunity

Retail shrink costs US retailers ~$100B+ annually; loss prevention software is a multi-billion dollar market with clear enterprise B2B ICP (multi-location retailers, convenience chains, grocers). No-hardware-replacement angle lowers friction for adoption. Monetization path is obvious via SaaS per-location pricing.

Medium Signal

Founder Signal

Liam Webster (CTO) is a Berkeley EECS grad (2023) with AI/CV research at ICSI and UC Berkeley security lab — technically credible but relatively junior with only internship-level industry experience outside research. David Elskamp founded a media agency at 15 that served Heineken and Adidas, showing hustle and sales ability, but his background is media/marketing not security tech. Together they're a reasonable technical+commercial pairing but neither has deep enterprise security or loss-prevention domain experience.

Low Signal

Competition

This is a crowded space with well-funded competitors: Verkada (multi-billion dollar valuation), Evolv Technology, Verint, March Networks, and AI-native entrants like Vantage AI and Aiplex all target retail video analytics and loss prevention. Major VMS players (Genetec, Milestone) are adding AI layers. No disclosed proprietary data moat or defensible differentiation beyond 'easier integration' is presented.

Medium Signal

Product

Live demo button, detailed feature descriptions (real-time detection, traceback, auto case building), and case studies with specific metrics (62% shrink reduction, $3k/month savings per grocery store, 60hrs saved weekly). However, no named customer logos, no pricing page, and the '1,000th shoplifter caught' milestone is a LinkedIn post claim without third-party verification. 'Request Access' gating suggests limited deployment scale.

OverallB Tier

Lexius has a real product with early customer traction (case studies citing specific savings, 1,000 shoplifters detected claim) in a massive, genuine market. The no-hardware-replacement positioning is smart for reducing sales friction. However, the competitive landscape is brutal — Verkada alone has hundreds of millions in revenue and does AI video analytics, and multiple well-funded AI-native competitors are targeting the same ICP. The founders are scrappy and technically capable but lack enterprise security sales experience and deep domain expertise, which matters enormously for selling into multi-location retail chains. Early metrics are promising but need to scale fast before larger players fully commoditize the AI layer.

Active Founders

David Elskamp

Founder

2x founder. Built a six-figure media agency in high-school.

Liam Webster

Founder

Lego kid → film lover → software explorer → seeing machines EECS B.S. @ UC Berkeley AI & Privacy Research @ ICSI Co-founder @ Lexius.ai

Lexius

TierB Tier

BatchWinter 2026

Team Size3

StatusActive

LocationSan Francisco, CA, USA