Graph Databases - Right tool for the right job

Graph Databases - Right tool for the right job

We are constantly in the state to strive for effectiveness and efficiency. And in that journey many a times, efficiency takes over, sometimes leading to sub-optimal decisions, especially when you refuse to take a step back, to look at alternatives and end up sticking to the traditional approach of maximizing the returns from your current assets and try to force fit solutions.

I have been reading about graph databases for a while and last year got an opportunity to position it for a challenging use case, where the traditional RDBMS approach was unable to holdup. POC helped establish the value of graph database for the specific use case with impressive outcomes. (Again… specific tool for specific purpose…..not a panacea). For the business use case where an earlier established solution using RDBMS columnar database approach took hours, the graph database spit the output in seconds…. Impressive indeed. Very excited about the graph database and more so considering the variety of the business use cases that it can help solve.

Some of the learnings:

• Graph database is especially suited for use cases where networks/connections are to be modelled…. The more the connections, the more effective the tool becomes…. which otherwise will be difficult to model using joins in RDBMS as the depth and the breadth of these connections is fluid.
• Graph database is just one component in the overall puzzle. You also need to solve for data ingestion, data transformation, modelling in graph db, querying and finally very important aspect…. the presentation of the output is a manner that’s easily absorbed by the consumer and prompts action.?
• Requires a different way of solutioning (thinking in patterns), modelling, querying. Unlike traditional SQL, the query language is specific to the platform.