OBSRVR

Your footfall data
is wrong.

Sensors double-count. They can't exclude staff. They can't track journeys. Your conversion rate is based on a lie. OBSRVR fixes that, using your existing CCTV cameras.

🎯 True footfall accuracy via computer vision
🔒 Zero PII stored; privacy by design
Seed Round · Confidential · May 2026
02
The Problem

Retailers are data-blind in their own stores

E-commerce tracks every click, scroll, and hover. Physical retail? Stores have hundreds of cameras recording 24/7, and do nothing with the footage beyond security.

  • No idea who walks in, how long they stay, or what they care about
  • Merchandising decisions made on gut feel, not data
  • Mall operators can't prove footfall quality to tenants or landlords
  • Existing solutions (RetailNext, Sensormatic) require ripping out cameras and installing proprietary hardware
0%
of CCTV footage used for business intelligence
$180B+
GCC retail market size, growing 6–8% annually. Vision 2030 accelerating KSA.
85%
of purchase decisions still happen in-store
03
The Accuracy Gap

Sensors lie. Decisions suffer.

What sensor-based providers get wrong

Double & triple counting
Beam sensors count every pass-through. A customer who browses, leaves, and re-enters = 3 visits. Footfall inflated 2–3x.
Staff counted as customers
Sensors can't distinguish employees from shoppers. 10 staff members walking the floor all day = hundreds of phantom visitors.
No demographic filtering
A luxury brand can't exclude children from footfall. A sensor counts a stroller the same as a buyer with intent.
No journey or dwell time
A beam at the door tells you someone entered. Not where they went, how long they stayed, or what they were interested in.

The downstream damage

Conversion rate with inflated footfall
2.1% "We need to fix our sales team"
Actual conversion rate (OBSRVR data)
6.8% "We need to optimize our floor layout"
04
The Solution

Turn every camera into a retail sensor

No new hardware. No personal data stored. On-device inference at the edge, so no cloud GPU costs and no raw data in transit. Just plug into the cameras already on the ceiling.

📷
Existing CCTV
Connects to any IP camera already installed
🖥
On-Prem Edge
CV processing on-site. Raw footage never leaves the store
🔒
Encrypted VPN
Only anonymized metadata transmitted via AES-256
OBSRVR Cloud
Aggregation, storage, report generation
📊
Dashboard
Real-time insights for staff and management
No face images stored. Only irreversible demographic metadata. PDPL/DDPL compliant.
05
The Moat

Every deployment gets smarter

Sensors are static after install. OBSRVR learns. A feedback loop where your team tunes the model means accuracy compounds over time.

1

OBSRVR detects events

The CV engine processes camera feeds and flags detected events: entries, demographics, group compositions, dwell patterns.

2

Your team grades accuracy

Ops staff review flagged events in the dashboard and mark them correct or incorrect. Simple thumbs-up/thumbs-down. Takes seconds.

3

The algo recalibrates

OBSRVR computes optimal parameters from graded samples and suggests adjustments. Your team accepts or reverts. The model gets sharper every cycle.

Accuracy over time

Representative trajectory based on internal testing

InstallWk 2Wk 4Wk 6Wk 8Wk 10Wk 12Wk 14
Sensor baseline (static)
OBSRVR (learning)

What this means

  • Competitors flatline after installation. Their accuracy on day 1 is the same as day 365.
  • OBSRVR accuracy compounds with every correction cycle. More usage = more data = better model.
  • After 90 days, the gap between OBSRVR and any competitor is wider than it was on day one.
  • Each client's model becomes an asset tuned to their specific environment.
In 90 days, we're more accurate than RetailNext will ever be in your store, because your team trained us.
06
Traction

$1.35M in contracted value

$299K Annual Recurring Revenue
$1.35M Total Contract Value 3–5 year terms
8 Contracts 6 closed, 1 PO pending, 1 in progress
3x Repeat buyer Flormar: 3 contracts, 18 branches
Status Client Scope Term ARR (USD) TCV (USD)
Closed Galleria Mall Full Mall 5 yr $163,324 $816,621
Closed MAF (Mall of the Emirates) Fashion Dome 5 yr $58,856 $294,278
In Progress Richemont Piaget, Chloé, MontBlanc: 10 branches 3 yr $35,967 $107,902
Closed Flormar 18 branches (3 contracts) 3 yr $35,902 $107,706
Closed Alaia 2 branches 5 yr $3,270 $16,349
Closed Panerai 1 branch 5 yr $1,635 $8,174
Total $298,954 $1,351,030
07
Products

Two products, one platform

Same CV engine, different value propositions for different buyers.

🛍

Retail Analytics

For retail brands and groups. Understand who walks into your store and how they behave.

  • Footfall counting & unique vs. repeat visitors
  • Demographics: age, gender, sentiment
  • Group types: individuals, couples, families
  • Dwell time & heat maps
  • Employee exclusion for accuracy
$17.50 / camera / month
🏬

Retail Leasing Intelligence

For mall operators. Prove footfall quality to tenants and optimize leasing strategy.

  • Zone-level footfall analytics
  • Tenant benchmarking & comparison
  • Peak hour & seasonal trend analysis
  • Demographic mix by zone
  • Leasing performance data for negotiations
$27 / camera / month
08
Business Model

SaaS with built-in lock-in

Per-camera monthly subscription

  • Retail Analytics: $17.50 / camera / month
  • Retail Leasing: $27 / camera / month
  • Contract terms: 3–5 years with auto-renewal
  • Expansion revenue: upsell across brands and branches
  • Zero hardware CAPEX for the client (uses existing CCTV)
Avg. contract value: $169K · Avg. term: 4 years
64–71%
Gross Margin
Cloud infrastructure only. No hardware COGS.
3–5 yr
Contract Length
Near-zero churn by design.
$25K
Monthly Recurring Revenue
Growing with pipeline. Dec peak: $35K MRR.
09
Why Now

A closing window

Three shifts are creating a first-mover advantage that won't last.

01

Competitors store PII. New laws ban it.

PDPL, DDPL, and ADGM data protection laws are arriving across the GCC. RetailNext and Sensormatic store video and facial data. OBSRVR stores none. Every new regulation widens our compliance moat.

02

Massive retail buildout needs data from day one

Vision 2030 is driving 100+ new malls and thousands of retail locations across KSA. These properties are being built right now, and every one needs analytics. First vendor in wins the multi-year contract.

03

Edge CV is finally cheap enough

What cost $500K five years ago runs on a $2K edge server today. This is what lets us price at $17.50/camera/month and still hold 64–71% gross margins. The cost curve is our pricing moat.

10
Market Opportunity

GCC retail intelligence market

Bottom-up sizing across mall operators, retail groups, and monobrands in UAE, KSA, Kuwait, and Qatar.

$400M
TAM
All GCC retail + mall accounts
$157M
SAM
Reachable within 3-year GTM
$6.4M
SOM
Target capture by EOY 2028
1,690
Total GCC accounts
215
Mall properties
1,453
Retail brands & groups
4
Markets (UAE, KSA, KW, QA)
11
Go-To-Market

Two ICPs, one land-and-expand playbook

🏬
Mall Operators
Avg. deal: $194K ARR
The pain: Tenants demand footfall proof. Leasing teams negotiate blind. Zone-level data doesn't exist.
Buyer: Mall Technology / Operations Director
Champions: Leasing team, Finance (rental yield optimization)
Entry point: RFP or direct outreach via industry events
Cycle: 3–6 months (POC → pilot zone → full mall)
Expand: Additional zones, then portfolio-wide rollout
Evidence: MAF started with Fashion Dome, now expanding. Galleria: full mall from day one.
🛍
Retail Groups & Brands
Avg. deal: $79K ARR
The pain: Conversion rates calculated on inflated footfall. Can't optimize merchandising or staffing. No customer journey data.
Buyer: Store Technology / Retail Operations lead
Champions: Store managers, Security (existing CCTV), Finance (conversion accuracy)
Entry point: Proof of accuracy gap → pilot 1–3 stores
Cycle: 3–6 months (pilot stores → brand-wide → group-wide)
Expand: Additional brands within the group
Evidence: Richemont started with Panerai (1 store), now expanding to Piaget, Chloé, MontBlanc (10 stores).

The accuracy wedge opens the door

The first conversation is always about footfall accuracy. Once we prove the gap between sensor data and real counts, the conversation expands to demographics, journey mapping, and dwell-time analytics. Accuracy is the wedge. Intelligence is the platform.

3–6mo
Sales cycle
100%
Close rate on POC
12
Competitive Advantage

Why OBSRVR wins

OBSRVR RetailNext Sensormatic (JCI)
Uses existing CCTV ✓ Yes Proprietary cameras Proprietary sensors
No PII stored ✓ By design Stores video Partial
GCC data law compliance ✓ PDPL/DDPL native US-centric US-centric
Hardware CAPEX for client $0 $50K–200K+ $30K–150K+
Deployment speed Days Weeks to months Weeks to months
Demographics & sentiment ✓ Full suite Limited Footfall only
GCC presence & support ✓ HQ in UAE No local team Via partners

The moat: Competitors force 6-figure hardware CAPEX and months of approval. OBSRVR plugs into cameras already on the ceiling: OPEX line item, not a board-level capital request. And our system learns from every deployment. Sensors are static. Our accuracy compounds. After 90 days, the gap between OBSRVR and any competitor is wider than it was on day one.

CV roadmap: Phase 1 (current): validated product-market fit using SAFR (RealNetworks), proving the model before building custom infrastructure. Phase 2 (funded): proprietary CV with human-in-the-loop feedback. Not just a licensing cost play. A learning system that makes every deployment an asset, not just a revenue line. Target: gross margins from 71% toward 85%+.

13
Team

Built by operators who know the region

OBSRVR was founded in Sharjah, UAE. The team combines deep technical expertise in computer vision and cloud infrastructure with direct retail industry relationships across the GCC.

ES
Elie Salame
CEO & Co-Founder
9 years in security technology and retail intelligence across the GCC. Previously led technology solutions at Securitas UAE for 5+ years, architecting and deploying large-scale video analytics across the region. Now also Senior Solutions Engineer at RealNetworks (SAFR), giving OBSRVR direct access to the CV platform powering its product.
PY
Patrick Younes
CTO & Co-Founder
Product and engineering leader with deep expertise in data analytics and enterprise technology. Previously led engineering at VUZ, building data-driven products in Dubai. Certified in Product Leadership and Product Management (Product School). Owns the OBSRVR platform architecture, edge deployment pipeline, and cloud infrastructure.
14
The Ask

Raising $1M seed to scale across the GCC

40%
25%
20%
15%
Engineering & Product $400K
Sales & GTM (KSA expansion) $250K
Infrastructure & Licensing $200K
Operations & Working Capital $150K
H2 2026

Consolidate UAE

Close pipeline. Scale MAF & Richemont. Hire 2 engineers, 1 sales lead.

H1 2027

Enter KSA

First KSA mall deployment. Target Arabian Centres, Red Sea Mall. Hire KSA country lead.

H2 2027

Scale to $1M ARR

10+ active clients. Prove KSA PMF. Begin Series A prep.

2028

Multi-market

Kuwait, Qatar. $4M+ ARR target. Series A raise.

$1M
ARR Target, 18 months
10+
Active Clients
2
Markets (UAE + KSA)
Series A
Next Milestone
15
OBSRVR

The intelligence layer
for physical spaces.

$1.35M contracted. Blue-chip logos. Privacy-first architecture.
Starting with retail. Ready to scale across the GCC.

📧 elie.salame@obsrvr.ae
🌐 obsrvr.ae