Struggling with big data and making sense? Need a custom domain specific search based application where Lucene or Elastic search out of the box does not suffice? How about a really intuitive and sleek visualization that pops out the information nuggets easily? Does your data require careful curation and normalization?
Sarvaha has been building for last 8 years semantically intelligent applications to analyze large heterogeneous datasets. Using NLP, Machine Learning models and Ontologies, we have solved complex enterprise problems. We have expertise in writing scalable search based applications that make use of distributed procesing frameworks such as Hadoop and NoSQL stores such as MongoDB along with search tools such as Lucene and ElasticSearch. We understand many open-source text mining and ML tools such as Stanford NLP, Apache Mahout and Jena. Most often the data requires careful curation and cleaning, a smartly designed feature extraction for acceptable performance and well thought-out visualization to leverage the inherent strength of the information. More importantly we bring the intuition to use the right method to solve a problem after understanding the domain requirements.
We have built automotive repair intelligence apps for a large enterprise that analyzes huge amount of historical repair data to provide predictive operational as well as strategic intelligence. We have helped a prestigious university in building an information retrieval system to aid cancer researchers.
Our current technology skills are listed below. However, our primary skill is Problem solving. And we learn whatever it takes to solve the problem.