From September 10–12, 2025, the 5th International Transport & Supply Chain Management Industry Exhibition (SCM FAIR 2025) was held at KINTEX, Goyang, Korea. SCM FAIR is a global event showcasing the latest technologies and trends across manufacturing, transportation, distribution, and logistics, all centered on supply chain management (SCM).
For the second consecutive year, Impactive AI participated with both an exhibition booth and a presentation at the SCM SUMMIT. At this year’s event, we had the privilege of meeting SCM professionals facing challenges such as overstocking and stockouts due to inaccurate demand forecasts—and sharing how Impactive AI’s technology can provide effective solutions.
In today’s fast-changing markets, unstable supply chains and limited resources constantly threaten the survival of businesses.
Forecasting technology helps companies effectively manage these risks. At the exhibition, many SCM professionals were eager to learn whether Impactive AI’s AI demand and price forecasting solution could address the challenges they face in the field.
Deepflow is a solution that automates complex demand forecasting processes, optimizes inventory management, and supports data-driven decision-making. By leveraging more than 200 advanced deep learning and machine learning models, it predicts product demand, raw material prices, and new product sales volumes.
With industry-leading accuracy, Deepflow has proven effective across sectors such as manufacturing, retail, and food. In practice, companies have reduced inventory costs by more than 30% and cut working hours by one-fifth.
At the booth, we shared customer success stories and explored tailored strategies with attendees from various industries. Impactive AI currently partners with leading companies such as Samsung Electronics, SK Telecom, CJ CheilJedang, Hanmi Science, and Dongkuk Industries, and held meetings with over 20 corporate representatives during the exhibition. Many visitors expressed strong interest by requesting follow-up consultations after their initial discussions.
One of the hottest topics among visitors was whether raw material prices can be predicted with high accuracy. Accurate forecasts enable companies to identify optimal purchasing timing and reduce costs.
Deepflow currently provides six-month forward-looking price trend forecasts for raw materials including metals (gold, iron ore, nickel, copper) and agricultural commodities (soybeans, cabbage, corn). By analyzing over 150 global market indicators, the solution achieves up to 98.6% accuracy in predicting raw material prices.
For commodities with publicly available price data, customized AI models can also be developed upon request. Many attendees engaged in extended discussions on raw material-specific forecasting possibilities, showing keen interest in how these capabilities could be applied to their own procurement strategies.
This year, we also introduced Deepflow’s new LLM-based analysis feature, designed to make complex forecasts easier to understand and apply in practice.
Even with highly accurate predictions, companies often struggle to act without understanding the rationale behind the numbers. The LLM-powered module addresses this by generating reports that include: Historical sales trends and seasonal patterns, Future demand outlook with supporting rationale, Department-specific action plans.
These insights empower companies to translate forecasts into actionable strategies quickly and effectively.
On the final day, September 12, Impactive AI’s Executive Director Tae-won Yoo delivered a 30-minute presentation at the SCM SUMMIT titled “Leveraging AI Forecasting Insights for Optimal Purchasing Decisions.”
The session explored challenges companies face due to external risks and traditional inventory management limitations, then presented Impactive AI’s forecasting technologies and case studies. Yoo also unveiled our research on the LLM Black Swan methodology, which helps sustain forecasting accuracy even in unpredictable situations.
When unpredictable events such as tariffs or wars occur, conventional forecasting models often struggle to incorporate them. However, as a company dedicated to forecasting technology, Impactive AI has devised a method using LLMs to address this challenge.
By leveraging the background knowledge inherent in LLMs, we measure the level of uncertainty of an event—what we call the Black Swan Index—and transform it into time-series data for modeling. A case study was also presented, demonstrating how the LLM Black Swan method was applied to predict raw material price trends.
Following the presentation, many visitors came to our booth to ask questions about the talk and to learn more about Impactive AI’s forecasting technology and solutions. We were deeply grateful for the overwhelming interest—so much so that people lined up in front of our booth to receive consultations.
While SCM FAIR has come to an end, we are preparing more opportunities to meet with those facing challenges such as overstocking, shortages, and ineffective raw material purchasing strategies.
From September 30 to October 2, 2025, Impactive AI and our solution Deepflow will be showcased at AI Festa 2025, held at COEX in Seoul. Live demos and consultations will be available, so please feel free to visit our booth anytime. If you book a consultation in advance, you’ll be able to meet with us without any wait time.
.png)
You can find Impactive AI at Booth B-17 in the AI Convergence Zone, Hall A on the 1st floor of COEX. Simply follow the rightmost aisle of the exhibition hall, and you’ll reach our booth.
Impactive AI is a specialist in demand and price forecasting. If you’re curious about how forecasting technology can improve your business performance, we invite you to meet with us for a consultation. We are committed to supporting data-driven, strategic decision-making for your success.
.svg)
.svg)
