Operational Forecasting
Without the Data Science Team
Kodaco delivers accurate, probabilistic time series forecasts via API or CSV export — at a fraction of the cost of building and maintaining your own pipeline. Use it as a managed daily pipeline, or call it on demand as an API.
Everything you need. Nothing you don't.
Built for SMEs that want data-driven decisions without hiring a data science team.
Timeseries as a Service
Post your daily history to one REST endpoint and get a tuned, probabilistic forecast back in minutes — with an e-mail when it is ready. Prefer files? Every forecast also downloads as CSV.
Hierarchical Forecasting
Reconcile forecasts at every level of your business — category, SKU, location — so aggregates and detail always add up.
Probabilistic Bounds
Don't just get a point estimate. Every forecast comes with configurable confidence intervals so you can plan for best and worst case.
Auto Machine Learning
Optuna-powered hyperparameter tuning selects the best model (LightGBM, Prophet, ensembles) via rolling cross-validation. No manual tuning.
Event & Holiday Awareness
Dutch & European public holidays, seasonal patterns, and custom events like Black Friday are automatically incorporated into every forecast.
Hosted in Europe
All data stays within EU borders. GDPR-compliant infrastructure you can trust for sensitive business data.
Per-Forecast Pricing
Pay only for what you use. No expensive data scientists, no infrastructure overhead — a fraction of the cost of in-house.
Drift Monitoring
Automatic alerts when model accuracy degrades over time, so your forecasts stay reliable as your business evolves.
How it works
From first contact to production-ready forecasts in four steps.
Get in Touch
Tell us what you want to forecast — volumes, demand, staffing, capacity — and how far ahead. We set up your account and API access.
Data Fit Assessment
Share a sample of your historical data. We check history length, seasonality, missing values and events to confirm your data supports accurate forecasts.
Train on Schedule or via API
Let Kodaco retrain and forecast every day on a scheduled run, or submit a series to the API whenever you need a fresh forecast. Model selection and tuning are automatic either way.
Consume via API or CSV
Retrieve probabilistic forecasts — median, upper and lower bounds — from the REST API or download them as CSV. Integrate into your planning tools in minutes.
Forecast in action
14 months of daily volume data and a two-month forecast. Hover over the marked event points to see what the model knows about them.
The solid line is the forecast, with the shaded band marking the 80% prediction bounds; the fainter line in the same colour shows the realised actuals.
Try it yourself
Create a free demo account and explore the dashboards our customers use, running on live public energy production data. No sales call, no credit card.
Forecasts
The latest daily forecast per series, with prediction bounds and CSV export.
Insights
Which features drive each forecast, from holidays to weekly patterns.
Drift
How forecasts compare with the actuals as they come in.
Optimisation
The cross-validation results behind every selected model.
Timeseries as a Service
Send your history, get a forecast. No pipeline to build, no models to tune.
1Submit your series
curl -X POST https://kodaco.dev/api/v1/forecast-jobs \
-H "Authorization: Bearer $TOKEN" \
-H "Content-Type: application/json" \
-d '{
"forecast_name": "store_12_sales",
"horizon_days": 28,
"series": [
{"date": "2026-01-01", "actual": 112.0},
{"date": "2026-01-02", "actual": 98.5},
…
]
}'Daily data, at least ~4 months of history. Returns 202 with a job id straight away.
2Get the forecast
GET /api/v1/forecast-jobs/1234
{
"job_id": 1234,
"status": "succeeded",
"forecast_name": "store_12_sales",
"forecast_id": "9781cd9ac16b2740",
"forecast_runs_url":
"/api/forecasts/runs/?forecast_id=9781cd9ac16b2740"
}You get an e-mail when it is done. Each forecast day comes with lower and upper bounds.
About Kodaco
Gijs KoendersFounder, Kodaco
Built by a forecasting practitioner
Kodaco was founded out of a simple observation: most SMEs have enough transactional data to forecast well, but don't have the internal data science capacity to turn that data into reliable operational forecasts.
We built Kodaco to close that gap — a production-grade forecasting pipeline that any developer or analyst can integrate with, no ML expertise required. Hierarchical models, automatic optimisation, probabilistic outputs, and event handling all come standard.
Infrastructure is hosted in Europe. Pricing is per forecast ID, so you only pay for what you actually use.
Let's talk
Want to see if Kodaco is a fit for your business? Drop us a message and we will start with a free data fit assessment.