SEOUL, South Korea – October 26, 2026 – Impactive AI (CEO Doohee Chung), a company specializing in AI-powered demand forecasting, announced today that it has overhauled the core features and user interface of its demand forecasting solution, Deepflow Forecast.
For companies managing hundreds or thousands of items, reviewing a demand forecast is only the first step. They still need to decide whether current inventory is sufficient and which items require order adjustments. Reflecting this real-world workflow, Impactive AI redesigned Deepflow Forecast so that users can identify demand shifts and inventory risks together, and move seamlessly from reviewing forecasts to cross-department discussion and final decisions.
At the center of the upgrade is an integrated demand and inventory dashboard that shows at a glance which items need attention first. The dashboard classifies each item’s inventory status as normal, short, excess, or out of stock, and automatically flags items whose demand has spiked or dropped sharply compared with the previous month or the same period last year. Managers can start with items facing inventory risks or sudden demand changes, without having to review forecast charts for every stock keeping unit (SKU).
Inventory management now accounts for item-level lead times. Users can check “Inventory Coverage,” which shows how long current stock can meet projected demand, and set shortage and excess thresholds based on actual supply lead times. For example, for an item that takes 30 days from order placement to being ready for sale, a shortage can be determined by whether current inventory can cover the next 30 days of expected demand.
New metrics also allow users to evaluate AI forecast accuracy and bias. By comparing past forecasts with actual results, users can see whether a particular item is repeatedly over- or under-forecast, and compare the AI’s forecast error with that of conventional methods, such as three- or six-month moving averages. The SKU detail page brings together historical shipments, forecasts, sales plans, current inventory, and projected inventory coverage, showing how forecast results affect actual inventory operations.
The solution also offers scenarios tailored to each company’s inventory strategy. Based on historical forecast error and bias, users can compare outcomes of applying the AI baseline forecast as is, adjusting quantities downward to reduce excess inventory, or adjusting them upward to prevent stockouts. This helps managers set order and inventory adjustments while weighing inventory costs against lost sales.
A new “Sales Anomaly Management” feature records the causes behind fluctuations in sales. Factors that numbers alone cannot explain, such as lost sales due to stockouts or temporary boosts from one-off promotions, can be logged as part of an item’s history. These records help users interpret past performance and, over time, build training data that reflects each company’s specific business context, laying the groundwork for more advanced forecasting models.
Deepflow Forecast also supports Sales and Operations Planning (S&OP), the process in which sales, supply chain management (SCM), and other teams jointly review forecasts and agree on a final demand plan. Teams can compare the sales team’s plans with AI forecasts and incorporate on-the-ground information, such as event schedules, to determine production and order quantities. Key agenda items, rationales, and final decisions that were previously scattered across departmental spreadsheets and emails are recorded in an integrated S&OP system and accumulated as a corporate data asset. This data will in turn be used to enhance future forecasting models, AI agents, and workflows.
An AI assistant makes this data easy to search and analyze. Users can ask about issues at a specific point in time or the causes of past demand changes and get the information they need. Through this feature, Impactive AI aims to reduce the time managers spend searching for information and reviewing forecasts, so they can focus on adjusting orders and inventory for the items that matter most.
“For demand forecasting to be useful in practice, it has to go beyond simply checking forecast numbers. Teams need to know which items to look at first and how to adjust inventory and orders,” said Doohee Chung, CEO of Impactive AI. “With Deepflow Forecast, we will connect forecasting, risk assessment, cross-department collaboration, and decision-making into a single workflow, and continue to raise the quality of forecasts and decisions based on each company’s accumulated data and field expertise.”
Founded in 2021, Impactive AI has established itself as a leader in AI forecasting by advancing its technologies for demand forecasting and raw material price forecasting. The company took part in Seoul Startup Hub’s Open Innovation program in 2024 and won the Grand Prize at IMPACT DIVE, hosted by the Seoul Business Agency (SBA), in recognition of its results. Impactive AI also develops innovative forecasting technologies, including a black swan data generation methodology that combines quantum computing-based forecasting models with generative AI, helping companies strengthen their response to supply chain risks.
About Impactive AI
Impactive AI is an AI solutions company specializing in product demand and raw material price forecasting. Its proprietary solution, Deepflow, forecasts product sales and raw material prices, and also recommends order quantities along with the inventory and purchasing issues that managers need to check each day. Managers simply review and approve the recommendations. This reduces repetitive, spreadsheet-based manual work while cutting inventory and raw material procurement costs. Managers’ decisions and their rationale are accumulated as a corporate data asset, continuously improving the quality of future decision-making.
Media ContactCompany Name: Impactive AIContact Person: Jeon Hyun-young, ManagerEmail: Send EmailCountry: South KoreaWebsite: https://www.impactive-ai.com
