Linguatools describes themselves as "experts in computational linguistics and human language technology, working" in the following areas:
- creating text corpora and dictionaries,
- implementing tools for linguistic text analysis and semantic similarity,
- training of SMT (e.g. Moses)systems
- integrating MT into the translation workflows
- using Apache UIMA, Apache OpenNLP, Okapi localization framework, Apache Lucene and SolR, Apache Nutch, Weka machine learning toolkit, Neo4j and more...
Here's a link to their corpora: http://linguatools.org/tools/corpora/