All projects
AI & ML
Consumer Goods
Meridian Forecast Engine
Demand forecasting that lifted accuracy 21 points.
Client
Meridian CPG
Delivered
2024 · 5 months
Team
7 specialists
Category
AI & ML
Overview
What we built
Replaced spreadsheet forecasting with a hierarchical ML pipeline that plans 8,400 SKUs across 26 countries.
Business impact
Outcomes that moved the P&L
+0 pts
Accuracy lift
0
SKUs planned
0%
Cycle time
₹0 Cr
Capital freed
The challenge
- Forecast accuracy stuck at 61% weighted MAPE despite six analysts.
- Plans took 9 days to produce and were stale on arrival.
- Stockouts and write-offs both trending in the wrong direction.
How we approached it
- Built a hierarchical forecast reconciled at SKU, region and channel.
- Combined statistical baselines with gradient boosting for uplift.
- Wired plans directly into S&OP and replenishment workflows.
What changed
- Forecast accuracy improved from 61% to 82% weighted MAPE.
- Cycle time dropped from 9 days to 6 hours.
- Working capital freed up: ₹115 Cr in the first year.
Project gallery
Inside the delivery
Hierarchical model
Reconciled SKU / region / channel forecasts.
S&OP workspace
One-click accept or override.
Backtesting studio
Confidence before every release.
Technology
Built with
The tools and platforms we chose for this engagement.
Python
Prophet
LightGBM
Snowflake
Airflow
Streamlit
Talk to our team
Have a similar challenge?
Tell us about your goals. We'll put together a scoped conversation with the right specialists.
Ready to amplify your data?
Book a 30-minute strategy call. We'll map your data landscape and identify the three highest-leverage moves.