Competence

Competence Areas

Artificial Intelligence

In previous projects, we have incorporated rules systems as engines used by other servers in a backend. We have even integrated machine learning algorithms in the rules engine, either inserting rules obtained from machine learning algorithms like C4.5 or adding adaptive “black box” functions whose outcome will be handled by the rules.

Machine Learning and Statistical Learning

Our experience includes statistical tests on biological and telecom data, as well as more advanced clustering, classification and regression algorithms, including C4.5, COBWEB, K-means or Support Vector Machines (SVM).

Data Mining and Social Network Analysis

We have applied data mining and graph analysis on different sets of data, such as telecom usage data. Past work has included both traditional, offline data mining as well as automating data mining as part of a process; and the setup of a social network server.

Products and Tools

Our languages of choice are C, Erlang and Java. However, we acknowledge that our client may have its own software development tools, so we are prepared to work in a plethora of other languages like Matlab, Lisp, Smalltalk or even C++.

We have experience with the following open source tools:

PENTAHO

Pentaho is an open source business intelligence solution covering all steps in a data analysis process: data preparation, modelling and analysis, and result deployment. Pentaho includes:

  • Pentaho Reporting – Access data and deliver information to the organization
  • Pentaho Analysis – Explore and analyze data interactively with rapid response
  • Pentaho Dashboards – Get immediate visibility into metrics and KPIs
  • Pentaho Data Integration – Cleanse and integrate data wherever it exists
  • Pentaho Data Mining – Discover hidden patterns and indicators of future performance
    • Our experience includes using the Pentaho Data Integration distribution tools to speed up the preparation of several Gbs of customer data, and extending Pentaho Data Mining with a distributed implementation of K-means.

      WEKA

      Weka is a collection of machine learning algorithms for data mining tasks. The algorithms can either be applied directly to a dataset or called from your own Java code. Weka contains tools for data pre-processing, classification, regression, clustering, association rules, and visualization. It is also well-suited for developing new machine learning schemes.

      JBOSS DROOLS

      Drools is a Business Logic integration platform that caters for the enterprise process and rule needs. It is composed of the following components:

    • Drools Guvnor: workflow and rule management system
    • Drools Expert: rule engine based on the Rete algorithm, as other rule engines like CLIPS and Jess
    • Drools Flow: workflow execution engine and development environment
    • Drools Fusion: complex event processing / data streaming engine

      APACHE SYNAPSE

      Synapse is Apache’s Enterprise Service Bus, designed for performance an efficiency. Synapse provides several ready-to-use protocols such as XML-RPC and FIX.