Talk to us
All solutions
All industries
All teams
How we work
Assess, build and run, with one accountable team before and after go-live.How we work 2 weeksfrom scattered ideas to a ranked, delivery-ready plan.The AI Assessment Data scientists and engineers shipping to production, then maintaining it.Data and AI engineering Your data stays yours, governed from the first day.Trust and governance
Featured stories
See all success stories
4.2 hto collect 2,800 financial reports with fifty agents in parallel, down from 210 hoursRead the story 35%more coverage of points of sale for Heineken field teamsRead the story 50k+cable designs tested virtually for CoficabRead the story 800K+calls a month transcribed, summarised and classified at NOSRead the story
Resources
Information Hygiene: The Security Discipline That Matters More Than EverRead the blog Artificial Intelligence Made in Portugal: ‘We grew more than 60% in one of the most competitive markets in the world,’ says DareData Partner.All news
Company
2019Founded in Lisbon by engineers. Today we build and run AI for large companies across Europe.About DareData 130+engineers, data scientists and principals in Lisbon, Porto and LondonMeet the team We are hiring engineers who want to build AI that runs.See open roles 50+customers served. Tell us about yours.Talk to us

AICEP ranks 30,000 companies by trajectory, and 90% of top picks grew

Thirty thousand companies, ranked by who will actually grow

Industry
Public Sector
Solution
Optimisation

Thirty thousand companies, ranked by who will actually grow

Industry
Public Sector
Solution
Optimisation

AICEP ranks 30,000 companies by trajectory,and 90% of top picks grew

Client
AICEP
Industry
Public Sector
Solution
Optimisation
Technology
Machine Learning
90%of top-ranked companies showed measurable export growth after deployment
60–70%precision at identifying high-potential companies, above baseline for comparable segmentation tasks at this portfolio scale
30,000companies in the portfolio, tiered by the model

AICEP supports a portfolio of 30,000 Portuguese companies, tiered by how much they already exported, which says nothing about which of them is about to grow. Companies are now ranked on trajectory, not volume, using five years of growth, digital readiness and their percentile position within their own sector and size cohort. 90% of the top-ranked companies went on to grow their exports.

With DareData, AICEP:

  • Ranks 30,000 companies on growth trajectory instead of export volume
  • Identifies high-potential companies with 60–70% precision, above baseline for comparable tasks
  • Surfaces hidden gems that volume-only rules overlook
  • Assigns every company to a high-touch or digital-touch service tier
  • Joins INE, Banco de Portugal and AICEP data in one monthly refresh

The challenge

Tiering by volume, not by growth

AICEP is Portugal's trade and investment agency, with a mandate to grow the country's export competitiveness. With 30,000 companies in its portfolio, the central question is where to invest relationship management effort.

Volume says nothing about growth

Companies were tiered with static export volume thresholds. That missed high-growth companies with modest absolute numbers, and said nothing about which ones were about to take off.

Prioritisation without a system

With a portfolio this large, prioritisation ran on geographic familiarity and relationship intuition. There was no systematic starting point for deciding who deserved more attention.

The picture split across three places

The data that predicts an export trajectory, from financial history to sector dynamics and digital maturity, sat with INE, Banco de Portugal and AICEP's own systems. Nothing joined them.

The solution

Rank on trajectory, not volume

Everything starts with one data foundation. A data lake on Google Cloud ingests, unifies and normalises inputs from INE, Banco de Portugal and AICEP’s operational systems, and refreshes them every month. Building it was a prerequisite for any modelling, and it stays useful beyond this project.

Segmentation replaces static volume thresholds with clustering. k-Nearest Neighbours groups companies by five-year export growth, digital readiness indicators and percentile-based benchmarks within their sector and size cohort, and elbow curve analysis sets the number of clusters. A dedicated hidden gem layer explicitly surfaces companies with strong relative trajectories that volume-only rules would miss.

Forecasting splits the portfolio into two problems, depending on the data available. For companies with no export history, content-based filtering compares their characteristics with successful exporters in comparable sector and size cohorts to estimate potential. For established exporters, a time series model forecasts growth from their own trajectory and sector context.

The results reach account managers through Looker dashboards, with no technical mediation. Adoption was earned rather than mandated: model outputs were translated into business language, managers could interrogate the logic against their own knowledge of each company, and early validated results gave them something concrete to trust. A model that could not explain itself would have been technically sound and operationally irrelevant.

How it works

From data to decisions, in 4 steps.

  1. 01Unify the dataA Google Cloud data lake joins INE, Banco de Portugal and AICEP's own systems, refreshed monthly.
  2. 02Cluster on trajectorykNN clustering groups companies by five-year growth, digital readiness and percentile benchmarks.
  3. 03Forecast by situationContent-based filtering scores first-time exporters, and time series forecast established ones.
  4. 04Tier the portfolioLooker dashboards give account managers a ranked portfolio and the reason behind each tier.

The impact

Effort where growth actually happens

  • 90%of top-ranked companies showed measurable export growth after deployment
  • 60–70%precision at identifying high-potential companies, above baseline for comparable segmentation tasks at this portfolio scale
  • 30,000companies in the portfolio, tiered by the model

90% of the top-ranked companies showed measurable export growth after the model was deployed. The model identifies high-potential companies with 60–70% precision, above baseline for comparable segmentation tasks at this portfolio scale.

The whole portfolio of 30,000 companies now runs on a dual-mode service model, high-touch and digital-touch, assigned by the model instead of by volume. Hidden gems, companies with strong growth that volume rules overlooked, are surfaced by design.

A monthly automated pipeline replaces the manual data curation behind each portfolio review. The data lake built for it is a lasting asset that serves more than this one model.

The solution behind this storyData and AI engineeringSee how we build it

Your AI story, next.

Tell us what you want AI to do for your business. We design, build and run it with your teams, and measure it like the stories above.

  • 01Use cases chosen for value, not hype
  • 02Systems built into your stack and data
  • 03Run, monitored and improved after go-live
Talk to usBook an assessment

More from production.

All success stories
Rui MendesHead of GrowthHave a similar problem? Book 30 minutes.No slides. We look at your use case together.Pick a time