Mobile Development
Artificial Intelligence
How forecast.ai turned scattered commodity data into an explainable forecasting platform — and compressed a week of guesswork into a single day of clarity.
forecast.ai was built to solve a problem that costs industrial businesses millions and rarely appears on a balance sheet: buying commodities at the wrong time. Their customers — mid-sized manufacturers across the DACH region sourcing aluminium, copper, and energy inputs — were making multi-month decisions on insufficient data. Sifars built the platform. It was acquired.
Commodity prices move without warning. A port disruption, a currency shift, a sentiment change three countries away — any of these can spike the price of aluminium or natural gas overnight. Yet the businesses most exposed were making procurement decisions on gut feel, scattered spreadsheets, and guesswork.
The data existed — scattered, unstructured, impossible to act on without specialists stitching it together. A procurement manager spent a week preparing a single sourcing recommendation. A poorly timed buy cost between 5% and 12% on that spend. For a business running €5M a year through raw materials, that is €250,000 to €600,000 in avoidable cost, year after year.
Sifars joined at proof-of-concept stage — a Christmas-deadline build to prove viability and secure funding. Three products were iterated over 18 months before the team landed on Commodity Desk. Each version revealed what procurement teams actually needed before they would trust a forecast with real money.
All commodity forecasts and data sources in one view. No assembling information before a decision can start.
Every forecast shows what is driving it — macro indicators, sentiment signals, supply-side factors. The why is always visible alongside the number.
Teams validate historical forecast accuracy before committing real spend to forward projections.
Price movements tracked as they happen — not after they have already influenced a decision.
Each customer sees only their commodities and their data. Clean separation, consistent trust.
Live forecasts and driver breakdowns accessible anywhere — not just at a desk when a decision cannot wait.
"The team stopped spending energy on 'is this the right dataset?' and started spending it on 'what do we actually do with what we're seeing?' That shift — from data wrangling to decision-making — was probably the most tangible change in how the whole operation felt day to day."
— forecast.aiCommodity Desk was acquired — the definitive signal that the platform had market value, not just market fit.
Before Commodity Desk, a procurement manager spent a week gathering inputs, building analysis, and preparing a recommendation. With the platform live, that cycle compressed to a single day. The forecast, the driver breakdown, and the back-test were already on the dashboard.
The bigger shift was in confidence. Teams stopped debating which numbers to trust and started deciding what to do with them. Data trust changed how every sourcing decision got made.
The platform reached 10 or more paying customers across the DACH region and was acquired by its original funder — proof of commercial traction, not just technical delivery.
If your business makes decisions on data that lives across different systems — and the gap between having information and acting on it keeps growing — you will recognise this case study. That gap is where significant cost quietly accumulates. Not on a single line. In every slightly wrong call, across every month of the year.
© Copyright 2026 Sifars – Mobile & Web App Development Company. All Rights Reserved Privacy Policy