Applied AI

Zarya

AI-powered freight document processing.

Overview

What Zarya does.

Extracts, validates and routes the paperwork behind every freight shipment — bills of lading, customs declarations, packing lists, certificates of origin. It reads any format, matches it against the shipment record, and flags discrepancies before they hold a container at port.

At a glance
Detail	Value
Stage	Seed
Category	Applied AI
Year	2026
Measured by	Documents processed per month, error-detection rate, time saved per shipment versus manual review.

An aerial view of a working container port — **Origin** Every container leaves with a stack of paper

Containers held at a customs terminal — **Delay** A discrepancy found at port is expensive

Why we invested

The thesis behind Zarya.

Freight documentation scales with trade volume, and Pakistan is a major origin for textiles, rice and surgical instruments. Every container leaving Karachi carries a stack of paper someone must read, compare and approve. Zarya automates the comparison.

Three years building document-extraction pipelines as a service for Gulf logistics firms. They productised the most repeatable part, and the first paying customers are the same firms that used to hire them by the hour.

Shipping and customs paperwork stacked on a desk — **Extraction** Read, matched and flagged before the shipment moves

Team

The people building it.

Karachi-based: a machine-learning engineer who built OCR for a Gulf bank, and a logistics operator who ran customs brokerage for one of Pakistan's largest freight forwarders.

Website zarya.ai
Stage Seed
Category Applied AI

A machine-learning engineer at a monitor — **Karachi** Document pipelines built for Gulf logistics

← Back to portfolio

Building something similar?

Building in applied ai? We want to hear about it.

Talk to us