M. J. Carrillo accurately modernizes cross-border customs operations with Cervo AI

M.J. Carrillo is a nearly 80-year-old, family-run customs brokerage operating along the U.S.-Mexico border, specializing in high-frequency, cross-border shipments. With deep regulatory expertise and a long-standing customer base, the firm has built its reputation on accuracy, reliability, and service.

But as tariff rules evolved and entry preparation became more complex, maintaining that standard required more than experience alone.

To adapt, M. J. Carrillo implemented Cervo AI.

The Challenge

Before Cervo, M. J. Carrillo’s entry preparation process was highly manual, even with some AI assistance:

• Documents arrived in multiple formats - PDFs, Excel files, XML - requiring constant downloading, organizing, and uploading

• Entry writers relied on macros, manual data entry, and repeated validation inside their ABI system

• Every step introduced opportunities for human error

At the same time, regulatory complexity was increasing rapidly.

New tariffs and trade actions - including Section 232, Section 301, and, more recently, Section 122 - added layers of logic that entry writers had to track in real time. In some cases, a single product required:

1. Breaking out value and weight across multiple countries of origin

2. Applying different tariff treatments to each portion

3. Recalculating entries as rules changed, sometimes within hours

Even when using a prior AI solution, the team faced limitations:

1. Limited visibility into how entries were generated

2. No direct control over how logic was implemented

3. Delays when requesting changes or handling edge cases

The issue was not just speed. It was a lack of control and transparency in an increasingly complex environment.

The Solution

Cervo AI gave M. J. Carrillo a more transparent and controllable workflow for entry preparation - without requiring them to replace their existing ABI.

Instead of receiving outputs after the fact, the team could now:

1. Review entries directly within Cervo before pushing to their ABI

2. See confidence levels and source data for each field

3. Identify and correct issues earlier in the process

“It just feels like you’re more in control,” Oscar said.

Several capabilities proved especially impactful:

• AI Agent Studio - enabled the team to define importer-specific rules and logic, themselves, without waiting on external implementation

• Dynamic Parts Library - allowed for mass updates to tariff data, making it easier to respond quickly to regulatory changes like Section 122

• Parallel processing workflows - enabled the team to upload and prepare multiple entries at once, keeping work moving while AI operated in the background

• Hands-on adoption support - Cervo’s team worked closely with M. J. Carrillo during onboarding, helping them navigate early complexity and ramp quickly

At first, the transition required adjustment.

“I almost said no in the first few days,” Oscar shared. “It felt complicated at the beginning.”

But with continued use and support, the system quickly became intuitive - and significantly more powerful than their prior workflow.

“Once you get it, it’s really helpful.”

The Results

Within the first month of going live, M. J. Carrillo saw strong adoption and operational improvements.

70%+ of monthly shipments processed through Cervo within four weeks of go-live

With their largest importer running 100% of entries through the platform

97.5%+ accuracy with greater transparency

Cervo consistently delivered high-quality entry preparation, while giving the team visibility into how each field was derived

“I can see where the data is coming from and the confidence level - I already know it’s going to be there.”

Instead of discovering issues downstream in their ABI, entry writers could now catch and correct edge cases earlier, improving both efficiency and confidence.

Faster response to tariff changes

Mass updates in the parts library allowed the team to prepare for regulatory changes ahead of time. For example, when Section 122 changes were introduced, M. J. Carrillo had updates ready before the rule went live, ensuring the first shipments processed were already compliant.

Reduced manual effort on complex entries

Tasks like Section 232 origin splits and calculations were now handled more efficiently, reducing the risk of human error.

More time for customer-facing work

Instead of focusing on repetitive data entry, the team could now spend more time:

1. Advising customers on tariff changes

2. Preparing for upcoming regulatory shifts

3. Building relationships and pursuing new business