Predictive analytics that warns you before it costs you.
Turn your sales, finance and operations history into forecasts for demand, cash flow, capacity and account risk. Our predictive analytics services assess data quality, test forecasts against past results and show uncertainty alongside each prediction, so your team can decide how to act.
- A free 5-day audit comes first
- We sign an NDA before we see anything
- Fixed price before you commit
- You own all the code
Demand forecast
Example
What will we sell next quarter?
- Read three years of orders from the ERP
- Forecast next quarter by product, with a range
- Found 4 products likely to run out in week 6
Draft purchase orders for the 4 products?
An example with made-up figures. The buttons send nothing.
How predictive analytics works.
We read your history
Orders, invoices, payments and production records from your ERP, CRM and finance system. Nothing new to collect.
It finds the pattern
Models learn how your business moves through the year and forecast what comes next, with a range that shows how sure they are.
You decide early
Warnings arrive in the dashboards you already read, weeks before the problem would have shown up in a report.
Predictive analytics use cases.
Start with the decision you guess at most. Each tab sets today's guesswork beside what a forecast, with its range, adds to it.
- Today
Stock runs out, or sits on the shelf
With a forecastDemand forecast by product and by quarter
Your sales history, seasons and open orders become a forecast for each product, with a range. Buyers see which lines will run short or pile up while there is still time to order differently.
By quarterand by product, with a range
Start a project - Today
Cash surprises arrive at month-end
With a forecastCash pressure points, weeks ahead
Receivables, payables and payment habits combine into a rolling cash forecast. Your finance team sees the tight weeks coming and can move a payment or chase a customer early.
Weeksof warning before a tight month
Start a project - Today
Good customers leave without warning
With a forecastAccounts at risk, flagged early
Falling order sizes, slower payments and fewer logins add up to a risk score for each account. Your account managers get a short list to call this week, with the reason for each name.
Earlywarning on accounts at risk
Start a project - Today
Capacity is planned on gut feel
With a forecastStaff and machines planned against the forecast
The model turns forecast demand into the hours, shifts and machine time it will need. Planners see the busy weeks ahead and can hire, schedule or subcontract before the queue builds.
Aheadof the busy weeks, not inside them
Start a project
How you stay in control.
A forecast is only useful if you trust it. These four rules are built into every model we deliver.
The forecast advises, people decide
Forecasts and warnings inform your team; they do not act on their own. Anything that orders, pays or changes a plan waits for a named person.
- You choose who acts on what
- Every warning shows its reason
- You set the thresholds for alerts
Your data stays yours
Your records stay in your own systems and cloud account wherever your policy requires it. We train the models on your data, for your business only.
- Runs inside your cloud account
- NDA, MSA and DPA signed
- Each person sees only what their role allows
In the dashboards you already read
Your managers keep the dashboards and reports they already open each week. The forecasts appear inside them, fed straight from your ERP, CRM and finance system.
- Fed from your ERP, CRM and finance system
- Every figure traced back to its records
- You own the models and the code
You see how accurate it is
Every forecast is checked against what actually happened. You see the hit rate on the same dashboard, so you know how much weight to give it.
- Forecast against actual, side by side
- Confidence range on every number
- Retrained as your business changes
Analytics project timeline.
Four stages, each with its own go or no-go. By week three you have seen the model forecast a period that already happened, using only older data, and compared it with the real figures.
01 · Week 1
Audit
We look at the decisions you plan for and the history you hold, and pick the three forecasts that would change the most decisions.
You end withA plan and a fixed price.
02 · Weeks 2 to 3
Trial
We forecast a period that has already happened, using only the data from before it, and compare the forecast with what really took place.
You end withProof it works, or a clear no.
03 · Weeks 4 to 8
Build
We connect the model to your live systems, put the forecast in your dashboards and set the alerts with your planners.
You end withA forecast your finance team has signed off.
04 · Weeks 9 to 10
Go live
Your team plans with it for a full cycle while the dashboard tracks forecast against actual.
You end withFigures to decide the next forecast.
After launch we check accuracy every month and retrain the models as your business changes.
Free AI readiness audit.
Judge the forecast on your own numbers. Pick stock, cash or customers, and learn which forecasts your history can already support.
What you hold when it ends
Free, 5 days
- A map of the decisions you plan for and the data behind them
- The three forecasts that would change the most decisions
- A fixed price and a timeline for the first one
The map of your decisions and data is yours to keep, whoever builds the models.
What our clients say.
From clients in the US, Portugal, New Zealand and the UK.
Predictive analytics questions.
Cost, your data and how far to trust a number about the future: what buyers ask first.
How do you decide whether our data is ready for forecasting?
We assess historical coverage, missing values, consistency and whether the records capture the outcome you want to predict. Forecasts should be tested against held-out history and a simple baseline. If the data is insufficient, the first step is a data collection or quality plan.
How much does predictive analytics cost?
The price depends on how many forecasts you want, how many systems hold the history and how much cleaning that history needs. The free 5-day audit ends with a fixed price for the first forecast, agreed before any building starts.
How much history do we need?
Usually what is already in your ERP and finance system. The audit tells you whether your records are enough for the forecast you want, and what to start keeping if they are not.
How accurate will the forecasts be?
We show you before you build: in the trial we forecast a period that already happened and compare it with what really took place. After launch, accuracy is on your dashboard every month.
Do we need a data team to use it?
No. The forecasts arrive in the dashboards your managers already read, in plain figures with a range. We look after the models.
Who owns the models once they are built?
You do, from day one. The forecasting models, the training data and the code that feeds your dashboards sit in your accounts under your name, with full documentation. If we stop working together, all of it stays with you at no extra charge.
How do we work with a team based in India?
Your first call is with a founder. Our hours overlap US and UK business hours by 3 to 4 hours, updates reach you in the tools you already use, and we answer within 4 business hours.
From the blog.
Plain-language notes from our engineers on this kind of work, written for the owner who signs the check.
Predictive AnalyticsCash Flow Forecasting with Machine Learning: A Step-by-Step GuideHow machine learning cash flow forecasting works, what it can and cannot predict, and eight practical steps to try it on your own finance data.Read the post
Predictive AnalyticsInventory Optimization with Predictive Analytics for DistributorsA six-step guide for distributors: use the sales history already in your ERP to carry less dead stock and run out of fast movers less often.Read the post
Predictive AnalyticsCustomer Churn Prediction: What It Takes and What Data You NeedHow customer churn prediction works, the data you already hold that powers it, five common myths and a checklist to see if you are ready.Read the post
Free project estimate
Start your analytics project.
Bring that decision to a 30-minute call with a founder. You'll find out whether your history can answer it, roughly what a forecast would cost and which one we would build first. If your data can't support it, we say so.
What you get back
- An indicative budget range
- A realistic timeline
- What moves the cost up or down
- A reply within 4 business hours
- Answered by a founder
- NDA on request, before you share details
Tell us what you need
A few lines is plenty. You get a number before any call.