Cross-border ops / Carrier systems / AI-native execution

Finlay Crawley builds the control layers behind international logistics.

I work where cross-border operations, carrier management, customs reality, and software execution meet. I design practical systems that make international mail and e-commerce logistics faster, clearer, and easier to control.

The goal is not a prettier dashboard. It is usable control: better shipment visibility, cleaner exception handling, sharper customs workflows, and operational intelligence teams can actually act on.

See initiatives Email Finlay

Operations console

Operator-builder profile

Active

Domain

International mail / E-commerce logistics

Built around real carrier, customs, and exception complexity.

Systems

Visibility / Workflows / Automation

Dashboards, manifest tooling, customs logic, and operational interfaces.

Working style

Fast iteration with domain context

AI speeds delivery, but the operating model stays grounded in reality.

Bias

Useful systems over software theatre

Interfaces should reduce ambiguity, not decorate it.

Execution loop

  1. Observe the operational failure mode
  2. Model the rules, states, and handoffs
  3. Prototype quickly with AI-assisted tooling
  4. Ship a system people can actually use

  • Focus

    Customs and carrier control

  • Builds

    Dashboards, workflow layers, automation

  • Methods

    Operational intelligence and exception design

  • Tooling

    Codex, Claude Code, scripts, and shipping logic

01 / Overview

Operator first. Builder by necessity.

Most logistics problems are not abstract. They show up as mismatched data, carrier handoffs, customs friction, broken exception paths, and teams working too hard around system gaps. That is the territory I care about.

I work in cross-border operations and carrier management across international mail and e-commerce logistics. That means dealing with the real operating model: shipment events, customs inputs, dispatch rules, routing constraints, exceptions, and the messy gaps between systems.

I build practical layers around that reality, from shipment visibility and exception handling to multi-carrier workflows, manifest support, customs and tariff automation, dashboards, and operational intelligence tools that help teams intervene earlier and with more confidence.

The throughline is simple: turn operational ambiguity into usable control. If a system does not make the work clearer, faster, safer, or more inspectable, it is not finished.

Cross-border operations Carrier management Customs workflows Tariff logic Shipment visibility Exception handling Operational intelligence Dashboards and reporting Workflow automation Python and APIs Codex Claude Code

02 / Systems

What I like to build around the problem.

The interesting work usually sits in the gaps between carrier reality, customs requirements, and the way teams actually operate. That is where good internal systems earn their keep.

A1

Operational intelligence

Monitoring layers that bring route signals, carrier performance, country changes, and disruption context into one decision surface.

  • Lane and country watchlists
  • Issue briefs and escalation context
  • Early warning for operational change

A2

Shipment visibility and exception control

Shared views that turn fragmented tracking feeds into milestone clarity, failure states, and action-ready queues.

  • Normalized event and milestone models
  • Actionable exception queues
  • Searchable shipment history and ownership

A3

Carrier workflow systems

Tools that reduce manual drag across manifesting, dispatch preparation, routing logic, and multi-carrier coordination.

  • Manifest shaping and validation
  • Carrier rule handling
  • Operational handoff support

A4

Customs and tariff automation

Rule-based support for customs data, tariff logic, declarations, and compliance-aware decision making.

  • Rules and decision support
  • Data quality checks and guardrails
  • Faster customs-ready workflow output

Visibility before velocity

Teams move faster when they can trust state, ownership, and failure context.

Edge cases are the product

The happy path is rarely the hard part in cross-border logistics. The system should respect the ugly paths.

Automation must stay inspectable

If operators cannot see the reasoning and state changes, the workflow becomes fragile.

03 / AI Workflow

AI is part of the operating model, not a side gimmick.

I use AI seriously in my workflow, especially Codex and Claude Code, to move from operational pain point to working system faster. The value is not empty generation. It is compression: tighter scoping, faster prototypes, quicker iteration, and more time spent testing the real edge cases.

When the domain is messy, context matters. I bring the operating model: carrier nuance, customs logic, failure states, data requirements, and the lived reality of how a team actually works. AI then becomes a force multiplier for design, coding, and iteration.

That combination is especially effective for internal tooling, dashboards, workflow layers, automation, and system glue that might otherwise sit in a backlog for months. It lets me test more ideas, surface better interfaces earlier, and tighten the build around real operational constraints.

This is not AI theatre. It is a faster path from operational knowledge to useful systems.

Build loop

  1. 01

    Map the operational problem

    Break the workflow into states, actors, rules, handoffs, and failure modes.

  2. 02

    Shape the system quickly

    Use Codex and Claude Code to explore interfaces, scripts, data models, and logic.

  3. 03

    Pressure test the edge cases

    Check customs differences, carrier quirks, exception states, and operator usability.

  4. 04

    Ship and refine

    Keep tightening until the tool reduces real operational load instead of adding another layer.

Codex

Repo-grounded implementation, interface iteration, code changes, and technical validation.

Claude Code

Decomposition, alternate solution paths, second passes, and sharper design iteration.

Operator context

Real workflow knowledge, domain rules, and human judgment over what actually deserves to ship.

04 / Projects

Selected initiatives and system themes.

These are the kinds of systems and initiatives I am focused on across logistics, customs, operational intelligence, and AI-assisted product building.

Showing 6 initiatives across operations, carrier workflows, customs, and AI-native building.

Operational intelligence

Global Intelligence

A global monitoring and briefing layer for routes, countries, and carrier performance. The aim is to catch disruption, change, and emerging risk early enough to act before it spreads downstream.

  • Route and country watchlists
  • Disruption summaries and escalation context
  • Operator-facing intelligence briefs
Ops intelligence AI-native build

Visibility system

Shipment Hub

A shared control surface for shipment visibility, milestone tracking, and exception workflows across multiple carrier feeds. Built to make fragmented tracking data operationally useful.

  • Normalized milestone model
  • Exception queues and intervention views
  • Searchable shipment history and state
Ops intelligence Carrier workflows

Carrier workflow

Multi-Carrier Manifest Tool

A manifesting layer for handling different carrier formats, dispatch rules, and downstream requirements without relying on brittle manual steps.

  • Carrier-specific manifest shaping
  • Validation before handoff
  • Cleaner dispatch preparation
Carrier workflows

Compliance workflow

Tariff and Customs Automation

A rules-first workflow for customs data preparation, tariff logic, and compliance support. Designed to reduce repetitive checking while keeping decisions inspectable.

  • Rule libraries and decision support
  • Data quality and exception handling
  • Customs-ready workflow output
Customs automation AI-native build

Product approach

AI-Native Workflow Building

An internal product approach that uses Codex and Claude Code to turn operational knowledge into working tools quickly, especially where teams need custom workflows more than generic software.

  • Specification to prototype loops
  • Fast UI and script iteration
  • Human review over every important decision
AI-native build

Exception handling

Exception Control Room

A triage workspace for live failures, backlog pressure, and root-cause discovery across carriers and lanes, built for fast diagnosis instead of passive reporting.

  • Priority views and ownership tracking
  • Pattern spotting across repeated failures
  • Faster escalation and recovery loops
Ops intelligence Carrier workflows

05 / Contact

Build something useful.

If you are working on cross-border operations, carrier workflows, customs tooling, shipment visibility, or AI-assisted internal products, I am happy to talk.

The easiest way in is email. The rest of my channels are below.

fin@finlaycrawley.com
Website finlaycrawley.com GitHub github.com/Fincr X x.com/finccr LinkedIn finlay-crawley-b46418140 Email fin@finlaycrawley.com