Kymos ©

Elif Lab's suite of instruments

Elif Lab uses Kymos ©, the proprietary suite, the result of its in-house research and development, to collect information and extract knowledge from heterogeneous data and sources.
Through artificial intelligence we make complex business processes more efficient, manageable and automatable.

Kymos Elif Lab

Tecnologie

Algorithms capable of delivering results, forecasting and making decisions based on past experience and capable of learning and adapting their behaviour to the data provided.
Deep neural network techniques to solve problems on images and texts. Deep neural networks allow to find solutions on issues that usually require human intervention and are difficult to solve with other tools.
Bayesian models to transform the expertise from the consultant's and industry expert's domain into artificial intelligence algorithms validated by data.
Bayesian models allow to operate on abstract concepts and are able to obtain results even from limited amounts of data.
Techniques of classical statistics to explore and evaluate data. The classical statistical tools are fundamental to bring out the potential and value of the data.

Features

Heterogeneous sources

We retrieve, harmonise, process and analyse data collected from different sources: from CRM and business softwares, to web data, from document archives to system logs, from directories of files and information transmitted via API, to IOT devices and hardware.

Structured and unstructured data

Kymos © is able to manage and extract information from structured and unstructured data.
We analyze texts and extract meaning, attitudes and value from different types of documents (reports, papers, tweets, Facebook comments, reviews, invoices, rules and judgments...).
We identify texts, patterns, people, objects, feelings in videos and images, to make the most of audio and visual information.

Human interaction

The processed information becomes available in various formats.
Not only reports and tables, but also flows and models for automation, interactive dashboards to analyze activities in real time and make strategic decisions and chatbots, which can be integrated on websites, to investigate the data through human interaction.

Depending on the needs of the project, we can release the algorithms directly into client's IT systems or onto edge nodes, including through microservice and containers architectures.