EMD Customs cuts entry processing time by 75% with Cervo AI

EMD Customs is a family-run customs brokerage handling global imports across air, ocean, and ground. With a lean team of three, the company manages dozens of importers across diverse regulatory environments - requiring speed, precision, and deep compliance expertise.

To scale without adding headcount, EMD turned to Cervo AI.

The Challenge

Before Cervo, EMD relied on manual processes to prepare customs entries in NetCHB:

1. Entries could take 30-40 minutes, end-to-end

2. There was no efficient way to:

• Bulk-update HS codes or APHIS templates

• Automatically update parts

3. Frequent disruptions from tariff changes (ie, IEEPA removal) invalidated entire sets of parts

Even with experimentation, using tools like ChatGPT or Gemini for HTS classification, the process remained fragmented:

• HTS databases were difficult to search using natural language

• Security concerns prevented the team from using sensitive data in general-purpose AI tools

The Solution

EMD implemented Cervo to streamline entry creation, HTS classification, and compliance workflows - without replacing their existing systems.

Key capabilities included:

• Fast-tracked entry processing: end-to-end, from document ingestion -> parts enrichment ->  transmission into NetCHB, reducing manual touchpoints and accelerating entry completion

• Dynamic parts intelligence: a live-updating parts library supporting bulk edits, auditing, and PGA and Section 232 handling

• AI agent studio: enabling importer-specific SOPs, rules, and workflows for consistent, scalable execution

• Regulatory intelligence: streamlining PGA requirements, smart tariff stacking, Chapter 99 recommendations for accurate, compliant filings

• Tariff simulator: calculating landed costs, validating tariff impacts, and auditing entries in real time

The Results

After implementing Cervo, EMD saw immediate gains.

75% reduction in processing time

While EMD saw as much as 75% reductions in processing time on a per entry basis, this didn’t account for efficiency gains in other areas like maintaining parts or simulating tariffs.

Near-perfect accuracy

EMD takes accuracy very, very seriously. Whether the team is preparing 7501s, simulating tariffs, or using tools in beta, they’ve found Cervo to be “extremely accurate,” with “no complaints.”

Scaled output without additional headcount

EMD found that they had freed up time for higher-cognition work and decision-making, enabling the 3-person team to handle growing volume efficiently.