The brand behind the award-winning Loka Chai Maker replaced its spreadsheets with Lumina - connecting Shopify, Amazon and its 3PLs to forecast every channel, raise supplier POs, move stock to where it's needed and track landed cost by market.
Consumer Goods
Kitchenware & chai blends
D2C & Amazon
Shopify, Amazon US / UK / CA, subscriptions
End-to-end
Demand, purchasing, BOM, transfers, landed cost, reporting
Spreadsheets
Google Sheets forecast model, Xero
Loka is reinventing how the world makes chai at home. Its Chai Maker - designed with Pentagram and winner of a 2025 Good Design Award - brews authentic masala chai on the hob with none of the boil-overs, alongside Loka's own masala chai blends. What started as a Kickstarter launch is now a fast-growing, multi-channel brand: Shopify across the UK, US and rest of world, a loyal subscription base, and Amazon in the US, with the UK and Canada added in 2026 alone and new blends on the way.
Behind the scenes, the supply chain is as involved as the product. Each Chai Maker is assembled from parts made by several suppliers, with long lead times. Stock travels by sea or air into warehouses in the UK and US, then on into Amazon. Subscriptions run alongside one-off orders, Amazon bundles draw down individual SKUs, and the chai blends are planned around tea and spice harvests in India, so production is committed long before the demand shows up.
Almost all of this lived in a set of complex spreadsheets maintained by Kevin Davies-Hunt, Loka's COO. Forecasting, purchase orders, shipments and deciding where to send stock were all done by hand. With stock-outs the biggest risk on a 90-day lead time, Loka needed a system that planned the way Kevin already did - but ran on live data.
Loka chose Lumina to replace its spreadsheets with one connected system - shaped around how Kevin already plans, and fed live by Shopify and Amazon.
Rather than a black-box statistical model, Lumina's forecast mirrors how Loka already plans: growth and seasonality by product and channel, with every number open to override. Amazon and D2C are forecast separately, so Kevin can drop in his Amazon agency's numbers without throwing off the rest of the plan.
“Using the Lumina MCP, I got a volume forecast for the next 12 months, broken down by SKU for US only... and it created this amazing Excel file which I can now pass onto Stord.” - Kevin Davies-Hunt
When Loka switched US 3PLs, Lumina had a direct integration with Stord live in time for the cutover - so planning never skipped a beat.
Because Lumina is built on a flexible platform, the system keeps evolving with Loka.
Kevin used to spend half to a full day every Monday updating the forecast model, tracking forecast accuracy and copying and pasting numbers from Amazon. Now he doesn't do any of that. Since going live, Loka plans from one connected system instead of spreadsheets updated by hand. Sales and stock from Shopify, Amazon and the 3PLs flow in automatically, so Kevin starts from the current position rather than rebuilding it. Purchase orders are created automatically, transfers are recommended against real weeks of cover, and every shipment carries its own freight and landed cost.
The effect reaches beyond the planning team. Loka's accountants now log into Lumina and use its AI reports for month-end costing, and Kevin builds reports in Claude that combine Lumina data with Shopify and customer-service data.
The approach reflects how Lumina works with growing brands: start from how the team already plans, connect the data, and keep building as the business adds products, channels and markets.