The Problem & Challenges being addressed

The last 5 years or so has seen the e-commerce industry growing exponentially across the globe. Being a new industry / focus area for air cargo, a lot of processes with respect to sharing of shipment information across stakeholders such as airlines, forwarders, customs agencies are yet to be standardized.

This is particularly true when it comes to cross-border ecommerce, where delays and hold-ups can occur at each stage of the shipment journey due to incomplete information, lack of supporting documents, etc.

What complicates matters further is the fact that the air cargo supply chain today works in a sequential manner meaning stakeholders only pass information to the immediate upstream / downstream party.

Even if there's a clarification raised by a customs authority that needs to be acted by the shipper, this information today is passed from the customs authority to the airline, and then to the forwarder who reaches out to the shipper - which may easily take days (due to multiple emails, back &forth phone calls), thus increasing shipment idle time + handling/holding costs and negatively impacting end-customer experience. In many cases, smaller shippers may not be as familiar with regulations/technical terms which leads to even more delays to understand the nature of the issue and take corrective action.

Challenges addressed -

  • Challenge 1 - eCommerce Challenge: Develop a solution to e-commerce & air cargo, for example in the domain of customs & border, detecting undeclared dangerous goods, Customer Experience, Package Tracking, screening etc.

  • Challenge 2 - AI Challenge : Create an AI-powered tool or system that leverages IATA's ONE Record standard to enhance air cargo processes, optimization, efficiency, prediction detection etc.

The Solution

IBS Software's SkyCommerce is a one-stop platform for e-commerce shippers/sellers to share and resolve shipment record related information with various stakeholders (forwarders, airlines, customs authorities) in the supply chain.

The platform leverages AI to simplify the shipment data capture process for the seller/shipper, and can be integrated with any ONE Record server to share information, receive clarifications and resolve such issues with multiple air cargo supply chain stakeholders.

The SkyCommerce platform consists of 3 key capabilities -

  1. Smart data capture (powered by ChatGPT) where shippers/sellers can easily copy-paste order information that needs to be shared to stakeholders in a chat box or as an Excel upload

  2. ONE Record converter (powered by ChatGPT) where the above received order information is read and transformed into the JSON-LD structure of the ONE Record Create Logistics Object API in the backend

  3. Workflow Engine that receives clarifications, business errors received from multiple stakeholders via ONE Record and suggests corrective actions to the shipper/seller (upload a document, update relevant information, etc.) for resolution of the same.

How we built it

The SkyCommerce platform leverages ChatGPT APIs to extract order information from what's entered by the shipper/seller in the chat box or from the excel file uploaded by the shipper/seller.

The platform also uses ChatGPT to convert order information into a JSON-LD format that can be used to create Logistics Objects into the ONE Record server based on the ONE Record Data Model 3.0 and API model 2.0 published by IATA.

The front-end is on React, and the back-end is built on Java. We have integrated with a locally hosted NEONE ONE Record server in the back-end for creating and fetching Logistics Objects.

Challenges we ran into

The main challenge we faced was inconsistent connectivity with ChatGPT APIs over WiFi at the venue - which used to work / not work depending on the area and time of the day. This was circumvented by connecting to these APIs via our personal mobile data/internet hotspots.

Accomplishments that we're proud of

  1. We were able to use ChatGPT prompts to extract those attributes that are relevant for ONE Record from a chat interface. Once extracted, we were able to use ChatGPT prompts with examples to construct the JSON-LD request for the Create Logistics Object API based on these attributes. This was something that we did not expect to achieve without major issues.

  2. Since we used ChatGPT to extract and transform data into the ONE Record JSON-LD request structure, we were able to build the entire data capture and transformation logic with less than 50 lines of code. Using traditional coding, we would have needed atleast 5-10x lines of code to do the same.

What we learned

Video shooting and editing is no mere task!

What's next for SkyCommerce by IBS Software

We see the SkyCommerce platform (today designed for shippers to capture and resolve incorrect data ) as a tool that can be extended to other stakeholders such as forwarders and airlines as well. The workflow engine that today processes clarifications/errors received from other stakeholder systems would need to be configured with the business rules + context of a forwarder / airline in mind.

Built With

Share this project:

Updates