DEV Community

Ken Deng
Ken Deng

Posted on

The Matching Engine: Teaching AI Your Route, Cargo, and Service Logic

You've spent years building mental shortcuts—knowing which carrier to call for Shanghai-Rotterdam in peak season, or which rate to bump for dangerous goods. But when a new RFQ lands, you still manually cross-reference spreadsheets, cargo notes, and carrier preferences. The result? Delayed quotes and missed margins.

The solution isn't generic AI—it's a Matching Engine that encodes your unique decision logic into automated rules.

The Core Principle: Your Decision Logic as Code

Think of your expertise as a series of if-then-else statements. Your AI needs to replicate your thought process when you see a request: "Client needs 20 containers of temperature-sensitive pharma from Shanghai to Rotterdam in October."

Your brain immediately filters: Reefer carriers only. All-risk insurance required. Add 10% buffer for peak season congestion. Apply minimum markup. That's your matching logic. Write it down, structure it, and your AI can execute it in seconds.

The Framework: Route Priority Matrix

Build a Route Priority Matrix—a spreadsheet that scores your top 5 carriers (1–5) on Documentation, Communication, and Reliability. Then layer in Cargo Classification Rules: tag each rate library entry with suitable cargo types (e.g., "Reefer Only" for pharma). Finally, add Seasonal & Congestion Adjustments as rule tables.

For example, your rule for Shanghai-Rotterdam in September–November: Add 10% buffer to base ocean freight for congestion, or prioritize carriers with guaranteed space. For dangerous goods: Create a non-negotiable checklist—disqualify any carrier without DG certification.

Mini-Scenario in Action

A client requests a quote for high-value semiconductors (low weight, high value). Your Matching Engine automatically filters for carriers offering "All-Risk Insurance" inclusion, applies a cost-is-secondary priority, and generates a spot quote with your standard markup—all in under 30 seconds.

Implementation in 3 High-Level Steps

  1. Document Your Route Logic. On Day 1, list your top 10 routes. For each, write your 1st and 2nd choice carrier and the reason (e.g., "Maersk for reliability; MSC for cost"). This becomes your rule foundation.

  2. Synthesize into a Master Rule Table. On Day 4, combine your Route Logic, Service Logic (carrier scores), and Cargo Classification Rules into one spreadsheet. Include your markup strategies—like applying minimum 3–5% markup for commodity bulk, or premium margins for high-value goods.

  3. Test and Refine. On Day 5, input your most critical rule (e.g., DG handling) into your AI tool. Test it with a past RFQ. On Day 6, generate a quote for a new RFQ, compare it to your manual choice, and adjust one rule. Repeat until the AI matches your judgment 90% of the time.

Key Takeaways

  • Your competitive advantage isn't data—it's your decision logic. Codify it as rules.
  • Build a Route Priority Matrix and Cargo Classification Rules to automate carrier selection.
  • Layer Seasonal Adjustments and Markup Strategies to mirror your pricing psychology.
  • Test with past RFQs, refine one rule at a time, and scale by adding client-specific preferences from your CRM.

The goal isn't to replace your expertise—it's to let it run on autopilot while you focus on the exceptions that truly need your human touch.

Top comments (0)