I'm Nielson, I'm a software engineer at Biodentify, a Dutch startup. My main interests are software/data engineering, machine learning, bouldering and of course Lego.
Biodentify aims to reducing the (climate and landscape) impact of unnecessary oil and gas drilling. We achieve this by analyzing microbial communities in the ground using modern DNA sequencing techniques. Whose results are inputs for supervised machine learning models to produce mappings of underground resources with a higher resolution than conventional seismological techniques, resulting in less drilling.
After DNA analyses of the ground samples we are left with big datasets, which have a high degree of many-to-many relations, that need to be stored so we can mix and match samples from all kinds of places for the ML models. Neither RDBMS and NoSql seemed like the right solution due too the above mention datasets and some design requirements. We ended up trying graph databases and applying for a Neo4j Start Up license. So far it's a great solution!
Graph database technology has been really refreshing so far and I'm excited to become part of this community!
In my experience in sql e no-sql I see linear simplicity until you try to classify relationships.
The reality is complex, the graph-db can include more part of reality.
The nice thing is that Neo4j have progressive program for learning and a long-time support, as you can see better in the certification program : https://neo4j.com/graphacademy/neo4j-certification/.