Released in The Interline’s AI Report 2024, this executive interview with Retraced is one of a fourteen-part series that sees The Interline quiz executives from companies who have either introduced new AI solutions or added meaningful new AI capabilities into their existing platforms.
For more on artificial intelligence in fashion, download the full AI Report 2023 completely free of charge and ungated.
Key Takeaways:
- Retraced maintains user trust in AI-generated data through human oversight and continuous AI model training, ensuring all data undergoes a human review for accuracy and reliability before integration.
- By embedding AI directly into their platform where ESG data resides and leveraging secure Azure OpenAI data models, Retraced ensures seamless, secure access for ESG and CSR teams, fostering efficient and reliable transparency and compliance.
What’s your working definition of AI? Does it differ from the public understanding, which is currently dominated by large language models and generative text-to-image models? And how does that definition manifest itself in your solution(s)?
At Retraced, we see AI as technology taking on tasks traditionally handled by humans. In the past, these tasks exceeded the capabilities of available technology. However, with advancements in AI, we now can process and analyze vast amounts of data more effectively than ever before.
This shift allows AI to outperform humans in certain tasks. It all begins with extracting data from various sources, whether it’s from PDFs, images, or other documents. Previously, humans manually extracted this data, such as copying information from an audit report into another system. Then, the data needed to be carefully evaluated to derive actionable insights.
Today, AI streamlines this three-step process by seamlessly handling data extraction, evaluation, and even suggesting next steps.
While there’s understandably a lot of focus on how generative models might impact the creative side of fashion, there is still a huge amount of value to be extracted from using AI to automate tasks that are either too large-scale, too time-consuming, or too risk-prones and sensitive to be effectively done by people alone. How does Retraced AI apply this principle to supply chain visibility, supplier risk auditing, and compliance?
Our solutions support CSR managers and suppliers who deal with heaps of information. Sorting through all these files manually eats up a lot of time. Plus, it’s risky – spending hours on similar files can make you miss crucial details. But details are crucial for supply chain compliance. That’s where Retraced AI steps in: We extract data from all kinds of supply chain documents, interpret it, and provide actionable steps to enhance compliance. It’s about improving compliance and making a positive impact on the fashion industry simultaneously.
A significant barrier to AI adoption is building trust in the output. This is certainly true in places where generative models are being rolled out, but it’s especially true in an area as sensitive as supply chain transparency and accountability. When you have relationships between brands, suppliers, and other stakeholders that hinge on the accuracy and reliability of data – and that data is also the foundation for legal disclosures – how do you establish the right structures and engage the extended user base in a way that establishes real trust in AI?
We implemented what we call “keeping the human in the loop” for any data generated by AI. This means that before any AI-generated data is finalized and distributed, it undergoes a human review to ensure accuracy and reliability. This extra step is crucial for maintaining user trust.
While our goal is to eventually minimize the need for human intervention through continuous AI model training, currently, human oversight is essential. We ensure that no AI-generated data is integrated into our system without the opportunity for a final human check, reinforcing trust in our AI-driven solutions.
We’re also implementing service level objectives, meaning that we are setting ourselves AI accuracy goals. Continuous tests are conducted to verify that accuracy and ensure that we continuously improve.
Another key consideration is where AI lives in the enterprise technology stack. With off-the-shelf tools, AI resides separately from critical business data, and outside the reach of the established tools that ESG / CSR teams use. This is both a risk – auditing and compliance involve sensitive data – and a point of friction for the user experience. How does Retraced aim to both lighten the time and administrative burden of working with AI, and provide a secure way for users to centralise and analyse the information and the insights they need to build reliable transparency?
AI integration within the Retraced platform is strategic. We positioned AI exactly where ESG data resides, ensuring seamless accessibility for ESG and CSR teams. By embedding AI directly into our user interface, we empower users to leverage its capabilities effortlessly. We’ve opted for Azure OpenAI data models due to their robust enterprise grade data security standards to protect sensitive data from unauthorized use and exclusion from any training. This approach guarantees that sensitive information remains within our platform and isn’t utilized to train public models.
Looking beyond supplier auditing, what is Retraced’s vision for extending the use of AI into wider ESG and sustainability risk assessments? Where else do you see AI adding value in both responsive compliance with regulations and more proactive sustainability strategies? And do you believe the industry is ready to embrace AI in those areas?
We are already leveraging AI beyond supplier auditing. We use it for risk assessment and monitoring of suppliers, checking their compliance status and performance over time. AI support enables us to anticipate potential risks and supplier developments.
Our ultimate goal is to empower AI to provide forecasts, providing our customers with the chance to react proactively and leverage opportunities. To give an example: Imagine predicting the impact of climate change on the supply chain, allowing customers to be warned and giving them the chance to mitigate risks and look for alternative solutions in time.
I believe that the fashion and textile industry is more than ready for AI. Companies recognize its inevitability and embrace its potential. We’re receiving inquiries from potential customers eager to leverage AI technology for faster supply chain visibility and informed decision-making. They understand the urgency of making better decisions swiftly and are keen to explore AI solutions.
What do you see as the near-term future of AI – both within your solution(s) and in general? Do you believe it will be a transformative class of technologies the way people expect?
I’m genuinely optimistic about AI’s future impact. I believe it will significantly transform many industries. Picture a world where routine tasks are automated, allowing humans to focus on more meaningful work. This shift is already underway, with AI easing the workload for content creators and developers alike.
We believe in a future where AI enhances productivity and efficiency in everyone’s daily life. Our team is already fully embracing it, whether it is for creating concepts, designing our product or writing the code – AI enables our teams to drive innovation and achieve our goals more efficiently. The future holds promise, and AI is a key part of that journey.