Jaume Ferrarons Llagostera

Lead the R&D at SmartMonkey, to create AI and ML based products, capable of providing real-time decision making intelligence for last-mile operations. Main tasks include: - R&D of main optimization and ML algorithms. - Design production ML pipelines. - Lead product development. - Collect requirements from stakeholders and customers.

The main task was to investigate deep learning techniques on top of graphs to improve customer support processes by reducing search time. I was also involved in other projects to apply similar technologies on top of computer networks and data centers for performance prediction and root cause analysis. Some of the technologies used include: Tensoflow, Spark, HBase, MongoDB, R.

The purpose of the internship is to participate in a study and a Proof of Concept on predictive analytics based on Airport IT data to better predict the number of passenger entering an Airport. It would allow to optimize all the resources of an Airport based on this precise prediction.

In several databases containing publications appears the problem of synonymy and homonymy among the author’s name contained. During the internship, a system has been implemented to solve this kind of ambiguities in large-scale datasets efficiently using Data Mining techniques.

Developing scalable and real-time statistics and monitoring systems. In addition I developed parts of the web page and some hardware drivers. All using C++, C#, Java, PHP, MySQL.

I was a financial audit assistant. I developed programs and executed queries to obtain the required information for the audit. To perform the queries I used SQL, SAS and AS/2.

Developing the project LEARN-SQL: Learning Environment for Automatic Rating Notions of SQL. It is an automatic assessment system.