Managing 50,000 SKUs across multiple stores and warehouses requires more than spreadsheets and manual processes. In this session, Bragi Þór Antoníusson from BYKO shares how the company is transforming its supply chain through intelligent automation, shifting from warehouse-led planning to demand-driven replenishment. He explains how this new approach has improved stock availability, reduced inventory, and given teams more time to focus on customers, suppliers, and strategic decision making.
Enter your details to watch the recording on demand.
How BYKO Automated Replenishment Without Losing Human Control
Managing inventory across thousands of products, multiple warehouses, and retail locations requires more than good forecasting. It requires a clear operating model that allows people and technology to work together.
In this session, Bragi Þór Antoníusson explains how BYKO is redesigning its replenishment processes by automating routine inventory decisions while allowing people to focus on strategy, supplier collaboration, and customer needs. He explores the shift from warehouse-led purchasing to demand-driven replenishment, the importance of trusted inventory data, and why successful automation depends as much on organisational change as it does on software.
The presentation offers practical insights for retailers and distributors looking to reduce manual work, improve stock availability, and build more resilient supply chain operations.
Bragi Þór Antoníusson is Head of Supply Chain, Procurement and Marketing at BYKO, one of Iceland’s leading home improvement and building materials retailers. He leads the company’s supply chain transformation, focusing on automation, data-driven decision making, and building more efficient inventory and replenishment processes that improve both operational performance and customer experience.
Einar Thorhallsson: Welcome everyone. Today I want to talk about how we’re rebuilding AGR around two ideas: standardisation and AI assistance. The supply chain software market historically sold large, custom implementations. We think the next decade looks different.
Einar Thorhallsson: Most of our customers run an ERP — Business Central, NAV, SAP, NetSuite — and the ERP holds the truth about sales, stock and suppliers. AGR sits next to the ERP, reads that data, forecasts demand, proposes orders, and writes the approved orders back. No duplicate master data.
Einar Thorhallsson: On the AI side, the practical wins today are not autonomous agents. They are anomaly detection in sales history, smarter seasonality decomposition, and exception ranking so a planner managing 20,000 SKUs only looks at the 200 that actually changed this week.
Einar Thorhallsson: Automation is the layer underneath. Once a forecast is trusted, the order proposal runs automatically against supplier lead times, MOQs and pack sizes. The planner reviews, edits if needed, and approves. The PO syncs back to the ERP in seconds.
Einar Thorhallsson: On pricing and onboarding — we publish per-user pricing, and most customers are live in four to eight weeks. That’s only possible because we stopped building custom versions for each customer and standardised the data model.
Einar Thorhallsson: On results — across our customer base we typically see inventory holding down around 11% and stockouts down up to 40% in the first year. Those numbers come from our 2025 customer benchmark, happy to share the methodology after the session.
AI-driven inventory management uses machine-learning models to forecast demand at SKU level, detect anomalies in sales history, and propose replenishment orders that planners can approve in one click. In AGR, AI runs continuously on ERP sales and stock data and surfaces only the exceptions that need human attention.
Automation executes rules you already defined — for example, auto-generating a purchase order when stock hits a reorder point. AI improves the rules themselves: it learns seasonality, promotions and supplier lead-time variability from historical data and updates forecasts and safety stock without manual tuning.
Most AGR SaaS customers go live in 4–8 weeks. The standardised data model and prebuilt ERP connectors (Microsoft Dynamics 365 Business Central, NAV, SAP, NetSuite, Sage, IFS, Jeeves) remove the multi-month integration work typical of legacy supply-chain tools.
AGR has prebuilt, supported connectors for Microsoft Dynamics 365 Business Central, Dynamics NAV, SAP, NetSuite, Sage, Visma, IFS and Jeeves. Other ERPs connect via REST API or flat-file integration.
Across AGR’s customer base, companies typically reduce inventory holding by up to 11% and cut stockouts by up to 40% within the first 12 months — based on AGR customer benchmark data, 2025.
Yes. AGR is used by 400+ companies across wholesale distribution, specialty and FMCG retail, and manufacturing with raw-material and finished-goods planning. The same forecasting and ordering engine adapts to each vertical via configuration, not custom code.
Enter your details to watch the recording on demand.
Prefer to talk to someone? Request a demo
