The foundation · Data engineering platform

Your data, run by the people who know your business.

In fresh, everything changes. Now your team can change everything, fast and under control.

AI-powered · built on the latest frontier modelsyou approve

In use today

50,000+files brought in and processed, by email and upload
250,000emails handled
150,000report files produced
60,000data exchanges with ERP systems
350+ready-made steps to clean, map and combine data

Yesterday → today

Your data arrives from every side of the chain

  • Yesterday: Every retailer, supplier and ERP file in its own format

    Today: Brought in and mapped once, automatically

  • Yesterday: A new format means a ticket and a wait

    Today: Your team changes it the same day

  • Yesterday: The numbers live in one person's spreadsheets

    Today: The rules live in the platform, for everyone

Four promises

Data engineering your own team can run

Planners know what a calibre change or a retailer's new depot means. The data engineering platform gives them the latest frontier AI models to act on it.

01 · Change fast

Changes the same day, not a project.

A new retailer format, supplier or product, handled by your team the day it arrives.

02 · Change it yourself

A data team is welcome, not required.

Your planners describe the change in plain language and the AI drafts it. No code, no ticket.

03 · Change under control

A safety net on every change.

AI drafts, you check, and nothing goes live without approval. Every change keeps its audit trail: who, what, when and why.

04 · Keep what you learn

When someone leaves, the knowledge stays.

Not sure how a file or a plan is set up? Ask the assistant. It answers from the live setup, so the answer is never out of date.

A good packhouse team resets the line for a new retailer spec in an hour, without calling the machine supplier. The data engineering platform lets your team do the same with its data.

Where it fits

Automate data from along the chain

Brought in once, it feeds every plan.

From along the chain

  • RetailersEPOS, orders, forecasts
  • Suppliers and wholesalersplans, shipments
  • ERPsales, stock, orders
  • Excelthe sheets your team keeps

Data engineering platform

AI draftsyou approve

  • brings in the files and sources you choose
  • cleans, maps and combines them, without code
  • sets up planning, S&OP and Pricing
  • runs on a schedule, or the moment data arrives

Out to plans and systems

See it happen

AI drafts. You check. Nothing goes live without approval.

A retailer changes its file. Nobody calls IT: you ask the assistant, check what it flags, and the new numbers are in your Control Tower and your plan.

Retailer North changes its order file
no IT ticketAI drafts, you checknothing goes live without approval

Conceptual product view.

  • The right rights

    Each person can view, propose or approve, area by area, for as long as you decide.

  • Tried, then approved

    Every change runs on your real data first. Someone else approves it, with their name and reason on it.

  • Your systems stay yours

    Your ERP is only read, unless you agree that approved figures go back.

AI sees only what the task needs

Choose from a selection of frontier models, run through their public API or privately in AWS Bedrock. Setting up a file, the AI reads the layout with names masked, never your quantities or prices. Your data is never used to train AI models, and every request is logged.

How we hold your data

What it runs for you

Set up once, then it runs on schedule

Your ERP and Fresh4cast, in sync every night

Sales, stock and orders come in every night, or on the schedule you choose. Once the plan is agreed, its figures go back to your ERP, where you allow it.

The Monday availability report

One email at 07:00: this week against last, the movers at the top.

supplier emailsreport · 07:00

An alert when an estimate drops

A supplier's week falls past your line, and the buyer hears the same morning.

new estimatealert · on arrival

Retailer forecasts into the plan

Each retailer's file, in its own format, straight into the demand plan.

retailer filesdemand plan

One report, in each team's language

Spanish for Spain, Italian for Italy, the same numbers underneath.

weekly salesreport · translated

How a project works

We learn your business with you. At the end, your team owns it.

We spend the first weeks learning your data, processes and people alongside your team. It is about people and process, not software: a handover of capability, not a delay.

  1. 1

    We learn your business together

    Our data experts spend the first weeks with your team, learning your data, your processes and your people.

    your data, mapped
  2. 2

    We set it up with you

    Experts and AI bring in your data and set up S&OP and Pricing, and your team checks every rule.

    every rule checked
  3. 3

    Your team takes over

    We train the people who will run it day to day. From then on, they make the changes.

    your people, trained
  4. 4

    We stay behind you

    An accountable Fresh4cast expert approves sensitive changes and steps in when something unusual comes up.

    your team owns it

In fresh, everything changes. Now your team can change everything.

Tell us which files and changes slow you down today, and we will show you your team making them, fast and under control.