Book a demo
Book a demo

Matilde: Reduce the Time-to-Market of new products and optimize existing formulations.

The Explainable AI Platform that Accelerates the Development of New Formulations in Cosmetics, Chemicals and Food.

MATILDE is a platform that integrates predictive and generative AI to support R&D laboratories in the simulation, optimization, and validation of formulations — reducing time, costs, and the number of physical experiments.

Discover the right solution for your R&D

CONTACT US

Matilde optimizes:

first

MATILDE

after

Time to Market

7

Months

2
Months

Waste

50%

Material Waste
20%
Material Waste

Test

30

Testing before approval

10
Testing before approval

MATILDE: main modules

Monitor Laboratory data and manage R&D projects.

Predict, analyse and decide what to move forward in the laboratory

Verify regulatory constraints and possible alternative raw materials

FORMULATION MANAGEMENT
SIMULATOR
& EXPLAINER
SMART
ASSISTANT

How Matilde works

STEP

1

LEARNS

Uploads historical product data, tests and specifications

STEP

2

ANALYZES
Predicts performance and critical issues

STEP

3

EXPLAINS
Identifies key factors with transparent explanations

STEP

4

OPTIMIZES
Provides data-driven suggestions for the final formulation

Areas of application

COSMETICS & BEAUTY
AI for cosmetic formulation

Skincare, haircare, make-up, fragrances...
Predicts: viscosity, pH, stability, texture, absorption


Learn more
CHEMICALS & COMPOUND
AI for chemical formulation

Rubber compounds, paints, adhesives, polymers, additives...
Predicts: density, hardness, resistance, viscosity, elasticity, aging

Learn more
FOOD & BEVERAGE
AI for food formulation

Snacks and confectionery, supplements, baked goods...
Predicts: taste, stability, shelf-life, hardness, preservation

Learn more

Who is Matilde for?

Scientist & R&D Manager

Develop new recipes and optimize formulations

Product Developer & Innovation Team

Launch new products in response to market trends

Process Engineer & Quality Assurance

Ensure scalability and regulatory compliance

Technical Director & Plant Manager

Optimize costs and R&D resource allocation

Case Studies

FAQ

1. What is Matilde?

Matilde is the Explainable AI (XAI) and generative AI platform developed by Intellico for Research & Development (R&D) laboratories in the cosmetics, chemicals and food sectors. It optimises formulation, experimentation and testing processes, reducing the time-to-market for new products from 7 to 2 months.

Matilde predicts the chemical and physical properties of a formulation before it reaches the lab, reducing material wastage from 50 per cent to 20 per cent and the number of tests required before approval from 30 to 10. It also enables rapid access to documents, technical data sheets and regulatory specifications, as well as comparative analysis of formulations and ingredients. Matilde is available on the Microsoft Marketplace.

Matilde works through a four-stage workflow – Learn, Analyse, Explain, Optimise:
  1. Learn (loading historical data): Matilde connects to existing LIMS (e.g. Labware, Labvantage, Thermo Scientific), PLM and ERP systems, importing past formulations, laboratory tests and measured properties. Its Small Language Models (SLM) libraries also enable rapid consultation of information derived from technical documentation.
  2. Analyse (AI learning): Explainable AI models based on Graph Neural Networks (GNN) analyse the relationships between ingredients and identify hidden patterns and correlationsthat a formulator could not detect by hand.
  3. Explain: Matilde identifies the key factors determining the outcome with transparent explanations, so the formulator understands the reasoning behind every prediction.
  4. Optimise (prediction and simulation): once a set of ingredients is entered, Matilde returns data-driven suggestions for the final formulation.

Unlike traditional R&D software (PLM, LIMS), Matilde is natively based on Artificial Intelligence models that generate predictions, comparative analyses and similarity assessments from laboratory data.

Matilde does not replace existing systems, it complements them:

  • it works alongside LIMS, which collect and manage laboratory test data;
  • it integrates with PLM, which governs the product lifecycle following formula approval.

The key difference is where Matilde sits in the workflow. It acts earlier, during the most exploratory and critical phase of formulation: the formulator’s workbench, where they can track intermediate trials, explore alternative ingredients, compare similar formulations and validate hypotheses, all the way to the optimal formula. Matilde doesn’t just manage data, it turns it into actionable insights and predictions that support R&D decisions.

Yes. Matilde is designed to safeguard the confidentiality, ownership and integrity of formulations, which are key competitive assets. Organisations can use AI to accelerate innovation without giving up control of strategic data.

This protection is based on four pillars:

  • Full data isolation: customer data — formulations, tests and derived models — is kept entirely separate and cannot be accessed by other tenants or organisations.
  • Models remain company assets: models trained on customer data remain the exclusive property of the company and are not incorporated into shared models or used to train external models.
  • No third-party sharing: data is not shared with external partners or suppliers, nor used for any purpose beyond what has been agreed.
  • GDPR compliance: Matilde runs in Europe and complies with the GDPR (EU Regulation 2016/679) and security best practices, with access controls and policies in place to protect sensitive information.

Matilde is used by R&D laboratories across three industrial sectors:

  • Cosmetics & Beauty, for the development and optimisation of cosmetic and personal care formulations;
  • Chemicals & Compounds: chemical compounds, paints and plastic blends, with predictions of density, hardness, tensile strength and aging behaviour;
  • Food & Beverage: for optimising the composition, stability and shelf life of food products, including ingredient substitution within nutritional and regulatory constraints.

Matilde is designed for Research & Development and production teams:

  • Scientists & R&D Managers developing new recipes and optimising formulations;
  • Product Developers & Innovation Teams bringing new products to market in response to demand;
  • Process Engineers & Quality Assurance ensuring scalability and regulatory compliance;
  • Technical Directors & Plant Managers managing costs and the allocation of R&D resources.
Matilde integrates with the main systems already in use across laboratories and companies:
  • LIMS (Laboratory Information Management System) for example, Labware, Labvantage and Thermo Scientific;
  • PLM (Product Lifecycle Management);
  • ERP (Enterprise Resource Planning).
Through these integrations, Matilde imports historical formulations, laboratory test results and measured properties, turning years of accumulated data into a proprietary knowledge base that feeds its predictions. The company’s know-how stays with the company, even as people come and go.

Explainable AI (XAI) is the branch of AI thatmakes a model’s decisions transparent and understandable, moving past the black-box logic of conventional systems.

This is crucial because the user not only receives a prediction (e.g. the expected viscosity), but also understands which ingredients and properties drove it. This allows users to trust the result, learn from the process and make better decisions in a repeatable way. It is the principle on which Matilde is built on: “you decide, the AI explains why”.

Innovate faster with MATILDE:
available on the Microsoft Marketplace

Talk to one of our experts

Contact us for an evaluation of the impact of our Solutions on your Business.

Work with us

Intellico.ai is an Artificial Intelligence software company operating in Italy and across Europe.

Discover open positions