Using structured data and AI models to simplify Cross-Border Logistics.

When creating new digital software, whether you need a CRM (Customer Relationship Management), a TMS (Transportation Management System), or any other kind of software, it's one thing to make functional software that does whatever you need. Still, another challenge is creating software with the proper data structure to transform it into actionable business decisions. In this Experts Take, we went deep into how our Data team, led by Chloe Wang, Head of Data at Nuvocargo, is using structured data models to generate actionable automation and processes within Nuvocargo that generate savings in time, administrative work, and substantial costs for our customers and partners. 

NuvoOS, our proprietary TMS, launched earlier this year, giving our data team the main challenge of manipulating, using, and consuming the new set of data generated by the TMS. This required us first to define the data models' structure to find the most efficiencies and optimization within our data. 

NuvoOS’ Data Structure 

• Ingestion: Feeding NuvoOS native data into our models. 
• Storage: We separate variables by defined categories, allowing our teams to examine data through as many filters as needed. 
• Transformation: This data is transformed through different models, algorithms, and tools to generate potential insights and make actionable business decisions.  
• Consumption: Generate a visualization of this transformed data for different teams to consume daily, optimizing existing workflows, SOPs, and communications within different stakeholders. 

Having this data ecosystem and structure from the beginning allows Nuvocargo to create as many data products as possible with the benefit of having proprietary data.  

Integrations & AI implementation

To handle the integration of different data sources and streams within the transportation ecosystem generated by NuvoOS, Chloe Wang explains that we harness a robust ETL (Extract, transform & load) pipeline to ingest real-time data into our Snowflake warehouse. This gives us the support needed to generate an end-to-end view of our operations and allows our data analytics teams to perform complex queries and create dashboards to find real-time insights.

The aforementioned data structure will enable us to scale dynamically and manage increasing volumes of data efficiently, ensuring our decision-making is based on the most current state of events and information. 

With this level of structure, data streams that flow into each other and are warehoused in a way that can facilitate transformation and consumption, we have a prime ecosystem to add AI and machine learning models to our ecosystem, which enhances the efficiency and effectiveness of existing cross-border logistics operations. 

At Nuvocargo, we extensively leverage machine learning and AI within our proprietary platform, NuvoOS, to streamline cross-border logistics and enhance operational efficiency. Let’s explore an example of how we capitalize on structured data models that make a difference in our customers’ operations.

Route Optimization Project

To illustrate the effectiveness of our NuvoOs platform paired with our skilled team, we present a project that significantly enhances operational efficiency for one of our major clients—resulting in savings of approximately 1 million USD. This customer approached us intending to optimize their cross-border logistics operations to reduce costs. Our team at Nuvocargo eagerly accepted the challenge, using our advanced technology to deliver tangible outcomes.

We are able to forecast the most efficient routes and schedules by inputting and analyzing historical data—including traffic patterns, load requests, and driver performance—into our AI systems. This optimization has enhanced our client's operations, saving them 1.2 million USD a year by minimizing delays, cutting transportation costs, and ensuring the timely and cost-effective delivery of their goods.

Despite facing initial challenges with unreliable data in the first versions of this project, our model today is highly reliable. Our enhanced AI route optimization model now processes new requests in just 60 seconds, demonstrating our technological solutions' resilience and capability to overcome obstacles and deliver substantial benefits to our clients.

Real-time data & KPIs

At Nuvocargo, handling real-time data is critical due to the dynamic nature of logistics and cross-border transportation. Our multi-faceted approach leverages advanced technology to ensure timely and accurate data handling, crucial for making informed decisions. As mentioned, we employ a sophisticated event warehouse to capture real-time events within the NuvoOS platform. 

This enables us to track and store many data points as events occur, from vehicle locations and traffic updates to border crossing statuses and document processing times. This event-driven architecture ensures that all system components receive real-time updates, facilitating immediate responses to changing conditions on the ground.

For example, the enhanced accuracy brought by our most recent launch of Dynamic ETAs enhances visibility and gives shippers more control over what to do with their logistics operations. 

This is one of our newest digital products, but our team works tirelessly to find new efficiencies, automation, and solutions to simplify cross-border trade. 

When discussing key performance indicators (KPIs), we prioritize aligning with your needs. NuvoOS provides a wealth of reliable data; however, an overload of information can also hinder decision-making. To prevent this, we establish a Service Level Agreement (SLA) and Standard Operating Procedures (SOPs) with each new customer to clarify the following: 

• Which metrics are most crucial
• Which aspects of their operations are most important (such as cost versus delivery times)
• And other relevant factors

This approach ensures we provide the most essential data first. From there, we can uncover additional insights particularly relevant to business growth. 

In the words of Chloe Wang, our Head of Data, “The main challenge of having a tool or software like NuvoOS is that data is being generated at every moment. Having the discipline and vision to build a structured model that consistently generates reliable data for different teams to transform and consume creates efficiencies and AI algorithms that will only simplify the complexity of cross-border logistics today.”