Description

Kode is a technology partner that supports its customers in all phases of data analysis and exploitation, up to the deployment of Artificial Intelligence models. Kode develops custom projects, for which it relies on two proprietary frameworks: FactorAI for industry and Kodelivery for logistics. Kode also has on-shelf products such as: Alchemy and Vega toolkit (chemoinformatics and drug design), Kodemetrics (multivariate data analysis) and Spectralizer (spectroscopy and process control). All this is developed with Princess, a Data Science Framework that we use to quickly create any end-to-end AI-based solution.

The areas in which Kode operates are industry, mobility, logistics, sports analytics, chemical and pharmaceutical research.

Skills range from Machine Learning to Deep Learning, from Generative AI to Computer VIsion. Today, Kode has more than 20 talents, including computer scientists, chemists, mathematicians, statisticians, psychologists, engineers, graphic designers, with a background of multidisciplinary skills and experience.

As the world of Data Science is constantly evolving, Kode is also active in the scientific field, participating in congresses and industry events and publishing in international journals, and is always engaged in scientific research projects alongside public and private research organisations.

Services we offer

Kode produces custom projects and off-the-shelf products. Custom projects are applications developed according to specific customer needs. They are developed using, basically, a framework called FactorAI.

FactorAI is a complete system, consisting of a back end and a front end, into which are grafted both specific connectors for the customer's data sources and modules to implement the required tasks: e.g., OEE calculation, prediction of a maintenance intervention, quality control, and so on.

Kode has developed a number of vertical applications, which require only an initial configuration phase. These are: Spectralizer, that is a sw, which can be integrated into factorAI, that allows spectroscopic data to be processed and calibration models to be built that can then be used immediately in production. Kodemetrics, that is a web-based sw, very easy to use, that allows to run those basic algorithms for the work of every chemometrician: pca, doe, regressions and classifications, Janus, that is a small platform capable of calculating thousands of molecular descriptors and exploiting the potential of nearly 100 predictive models for calculating specific eco-toxicological properties.

Industries

Our applications are versatile and cross-cutting, so they are not restricted to a specific type of industry. On the other hand, as for experiences to date, these are in the following industries: chemical, pharmaceutical, food, tanning, and papermaking.

Reference projects

  • Predictive plant control for an chemical plant
  • Overall Equipment Effectiveness management for a papermaker
  • Anomaly detection system for a pharma company
  • Advanced process control management for a plant the makes plastic compound
  • A digital twin for predictive maintenance for a tannery
  • A color recognition system and recipe predictor for a dye factory
  • A production planning system for a fresh food producer
  • An automartic technical documentation reader for a chemical company.

Partner details