Artificial intelligence (AI) has been one of the defining tech trends of 2024 across almost all parts of our lives, but the software sector will be especially transformed by this technology. Indeed, according to one study by Tech Jury, more than a third of SaaS businesses (35 percent) are already using AI in their offerings, with a further 42 percent planning to incorporate it in the near future.
Another study by SaaS Academy notes that by the end of next year, AI will be integrated into nearly every new software product and service, and trade management software is no exception.
MIC's Data Science Team Lead Clemens Kriechbaumer, who has been leading the company's data analytics and AI efforts, has therefore been explaining what these trends can offer for trade management software and how MIC is looking to incorporate the technology into its offerings.
The key use cases for AI in trade management software
Some of the most intriguing use cases for this technology come in the form of generative AI (GenAI), which Clemens notes has become hugely popular in recent months. This has been driven by tools such as OpenAI's ChatGPT, which has helped boost public awareness of what AI can do.
While the underlying technology and architecture behind these systems have been around for a while, the power of Large Language Models (LLMs) now allows them to be fine-tuned to either take on specific tasks, such as customs tariff classification, or offer more general help to users.
Clemens said: “An interesting application is using such models as agents or assistants for various use cases in customs and global trade.” He noted that advances in GenAI will also be especially useful for streamlining search functions. This can enable software to move from traditional keyword-based searches to solutions that understand deeper semantic meanings.
This will allow for much better ways to interact with documents and data using retrieval patterns that incorporate LLMs as 'interpreters'. Clemens added that this is likely to become the standard way of searching sooner rather than later.
Looking at other applications, Clemens also stated: “It is conceivable to have AI-powered co-pilots for customs declarations, supporting the processing by providing suggestions and instructions for various fields.” However, he noted that in order to be successful, it will be important that these models and technologies remain available in an open-source manner and are not controlled by a small number of big tech companies like OpenAI/Microsoft or Google.
What are the challenges of adopting AI?
However, it is clear that AI is still a work in progress, and there are a number of challenges that will have to be overcome for the technology to be fully accepted, especially in regulated industries where the penalties for non-compliance are high.
For instance, Clemens observed that since GenAI is rooted in statistics, it can be challenging to use in compliance and legal tasks if there is no deterministic reasoning behind the outputs. Such details will be essential for explaining decision-making or results to authorities, so if an AI tool cannot justify why it has suggested a specific course of action, it may not be usable in some situations.
Similarly, there remains a lack of awareness about the actual capabilities of AI, which can lead to unrealistic expectations among users. Clemens stated: "Understanding of AI is still lacking among business users and decision-makers, making it easy to hype GenAI. The real challenge will be using GenAI to generate sustainable business value and products."
However, with the power of AI advancing rapidly, it will surely not be long before the promise of this technology starts to become reality. Check back soon to learn more about how MIC-CUST's data science team is adopting AI into our products.