Now Piloting · New York City

The intelligence layer
for textile recycling

LoomLogic's computer vision platform identifies, assesses, and classifies garments in real time, turning manual sorting stations into data-driven, high-margin operations.

Schedule a Demo See How It Works
97.4% Brand Accuracy
4.6s Per Garment
3.4× Value Lift
LOOMLOGIC VISION
CAM-01 · OVERHEAD · LIVE
LABEL NIKE · COTTON · GOOD CONDITION 97.4% ✓
QUEUE
+4 pending
CLASSIFICATION OUTPUT
BRAND NIKE
MATERIAL 100% Cotton
SIZE Medium (M)
CONDITION Good
CONFIDENCE 97.4%
ROUTE High Value
RECENT · 8 ITEMS
High Value Vintage Both None
ANALYSIS TIME 4.6s
Operator Dashboard

What's on the table,
quantified in real time.

Track performance across every garment, every shift, and every operator. LoomLogic gives your team the financial and operational visibility to make better decisions, faster.

T1 REVENUE · CFO PRIMARY
LIVE · ROLLING 30 DAYS
FLAGSHIP · REVENUE LIFT · ROLLING 30 DAYS
$4,659,171.00
▲ 46.8% vs prior 30d · +$355.3K · 46 stores · 0.2M items
WITH LOOMLOGIC HUMAN-ONLY BASELINE DAILY · LAST 30 DAYS
ANNUALIZED RUN-RATE
$56.70M
▲ 46.8%
Projected from trailing 90d
AVG LIFT / ITEM
$0.97
▲ $0.31 vs $0.66
STEP-CHANGE on Mar 4 cutover
ITEMS PROCESSED · 30D
5.2M
▲ +69.4% vs baseline
Across 46 active stores
VS HUMAN-ONLY BASELINE
+47.2%
▲ revenue per item
Paired A/B, n=12 stores
T1 CHANNEL MIX SHIFT CURRENT 30D VS PRE-DEPLOYMENT BASELINE
▲ +8.7 pts to premium tiers
WITH LOOMLOGIC
7.8%
14.9%
28.1%
33.0%
16.2%
BASELINE
38.4% DISCARDED
HIGH VALUE / VINTAGE
7.8%
RISE
14.3%
BETTER
28.1%
GOOD
33.4%
DISCARD
16.4%
T2 OPERATIONS · COO PRIMARY
T2 DAILY THROUGHPUT ▲ +69.4% post-rollout
ITEMS SORTED & PRICED · FLEET-WIDE
1.6k 1.2k 800 400 SYSTEM LIVE Feb 22 Mar 4 Mar 24
PRE740
→
POST1,254
PEAK1,547
T2 ITEMS / OPERATOR-HOUR TOP 5 STORES
BK-04 · Brooklyn
412 +124%
QN-02 · Queens
387 +102%
MN-06 · Manhattan
361 +80%
BX-01 · Bronx
339 +93%
BK-11 · Brooklyn
318 +83%
FLEET AVG 314 BASELINE 178 +76%
T3 GRADE ACCURACY ▲ +18.3 pts vs human-only
97.4%
80 PAIRED-STORE A/B 100
BRAND ID 97.4%
MATERIAL 94.1%
CONDITION 91.6%
INTER-STORE Σ ±1.8 pts
SAMPLE VIEW · NUMBERS PULLED FROM LIVE 46-STORE PILOT · UPDATED EVERY 60s v2.4.1 · API STATUS ●
Pilot Results

Validated in the field

Numbers from our live NYC demo sessions. Real garments, measured against human-labeled ground truth.

0
Garments Analyzed
Across NYC pilot demo sessions in testing
0
Brand ID Accuracy
vs. human-labeled ground truth
0
High-Value Capture Lift
vs. operator-only baseline
0
Analysis Per Garment
From placement to full classification

· Results from LoomLogic NYC pilot sessions, 2025–2026 ·

Process

From garment to insight
in seconds

LoomLogic integrates directly into your existing workflow. No manual intervention. No re-sorting.

01

Garment Intake

Items enter in any orientation or condition. No pre-sorting, no manual setup, no special handling required before analysis begins.

Garment Intake
02

Vision Capture

Our overhead 4K camera captures each garment under controlled lighting, eliminating shadow variance for consistent classification.

4K Overhead Camera
03

AI Classification

Deep learning models detect brand, material, size, and condition in real time. Full classification in under 5 seconds.

Real-Time Inference
04

Sortation Data

Classification results stream live to your operator dashboard and integrate with existing ERP or workflow systems in real time.

Live Dashboard
The Problem

A trillion-dollar industry
running on human hands

Textile recycling is one of the last major industries yet to be disrupted by AI. The result is massive economic and environmental waste at a scale that demands a technological solution.

92M
tons of textile waste generated globally every year
150B
garments are thrown away each year worldwide
7%
of all global landfill waste comes from clothing & textiles
Only 1%
of discarded textiles are recycled into new garments
20%
of discarded textiles are even collected for reuse or recycling
500,000+
global recyclers are still sorting textiles entirely by hand

The core bottleneck isn't collection. It's identification. Manual sorting is slow, inconsistent, and blind to value signals. High-value items get misrouted. Brands go undetected. Revenue is left on the table at every step of the way. LoomLogic fixes this at the source.

Our Solution

End-to-end AI
for sorting

We've built the full stack: interoperable hardware, state-of-the-art vision models, and a proprietary dataset, so that every garment that passes through gets identified, assessed, and routed correctly.

Computer Vision Software

State-of-the-art deep learning models trained on millions of labeled clothing instances, achieving real-time brand, material, and condition classification.

  • Real-time inference
  • Brand logo + label recognition
  • Material & fiber composition detection
  • Condition detection (Discard / Good / Better / Best / Wow)

Purpose-Built Hardware

Custom workstation hardware designed to ensure each garment is properly positioned and formatted for high-confidence classification.

  • Overhead camera array
  • Controlled lighting for precision
  • Low-latency edge processing
  • Retrofit-compatible with existing sorting tables

Proprietary Data Engine

A growing dataset of millions of labeled clothing instances, continuously improving model performance across every garment type and condition.

  • A rapidly growing proprietary dataset
  • Continuous data integration
  • Continuous model retraining
  • Improving accuracy over time
Purpose-Built Hardware

A workstation built
around the sorter's hands.

Purpose-built hardware that mounts directly onto any existing sorting table. No infrastructure overhaul. Plug-in intelligence that starts classifying from day one.

LoomLogic workstation hardware
4K Overhead Camera
Mounted arm captures each garment from directly above, with no blind spots or angle distortion.
Lighting Control
Controlled illumination eliminates shadow variance for consistent, high-confidence classification.
Operator Display
Live classification results surfaced directly at the workstation, in real time.
Retrofit-Ready
Installs on any standard high-top sorting table in under an hour.
Current Stage
Piloting with NYC Recyclers
Technology
Prototype → Production-Ready
Dataset
Millions of labeled instances
Why LoomLogic

Built for operators.
Designed for scale.

Whether you run a thrift store or a grading operation, LoomLogic delivers measurable value from day one.

For Recyclers & Operators

Stop leaving revenue
on the table.

LoomLogic automatically identifies and flags high-value garments your team would otherwise miss.

3.4×

high-value item capture lift
vs. manual baseline

Real-Time Brand ID

Identifies 30,000+ brand labels across garment types. Instantly, every time.

Lower Labor Costs

Reduce dependence on skilled manual sorters without disrupting throughput.

Live Operator Dashboard

Classification data and analytics streamed to your team in real time.

Retrofit-Ready

Integrates with your existing workspace — no new infrastructure required.

Currently Scheduling

Ready to see LoomLogic
in action?

We're actively onboarding pilot partners and scheduling demo calls. Fill out the form below and we'll reach out within 24 hours.

Book a 30-Min Demo Email Us Directly
or fill out the form

We'll reach out within 24 hours. No spam, ever.