Updated: Jun 10
What are the challenges that one can face while integrating Linkurious with other 3rd party tools?
Ans: The only skill required to successfully integrate Linkurious Enterprise is to know what the concept of APIs is and how to use them. Our embedded REST API server helps users to easily exchange information with Linkurious Enterprise and the exported (or imported) information is presented in a format that is easy to be parsed (JSON). Finally, REST API clients can facilitate the integration, they are widely diffused and available for almost all programming languages.
When we say personalization, does Linkurious re-use the report or graph that's built or does it create multiple copies based on the numbers of customization needed by multiple users?
Ans: With Linkurious enterprise , users can create different visualisations from using the data from graph DB.Linkurious Enterprise’s visualization can be customized by defining a set of styling rules. These rules are saved in the Linkurious Enterprise server and do not affect the data in the datasource: multiple users can save the same visualization multiple times (maybe with different styling rules) without affecting the datasource or the visualizations of the other users. There is also an option to apply default styles to the datasource so that all users can share common visual language.
Can I, as a user, add to the backend code to enhance the product, like a true federated model?
Ans: Either you can build your application with your own backend and use Ogma for graph interactions or you can extend Linkurious capabilities by using plugins which is a tailor made web application hosted on Linkurious server sharing the same authentication layer.
How does Linkurious raise alert in case of out of pattern behaviour, can I link it with mobile for paging?
Ans: Linkurious Enterprise helps you detect suspicious patterns in your data through Alerts which is a way to scan your graph database for specific patterns and display a match when such patterns appear in the data. A simple triage interface lets the users navigate through pattern matches and flag them as confirmed or dismissed. Alert detection queries are periodically executed on the graph database in order to detect even the most recent cases.The alert output using Rest API can be integrated to any communication server for Mobile paging or email alert.
What are the infra needs for the tool and can I integrate with couchbase, hive, hbase, CSV, mainframe etc?
Ans: Linkurious directly consumes the data from a graph database. Integration with table based tools can be done by creating a custom solution. For example : By using plugins or Custom API.
How do we ensure data localization is maintained if we are dealing with transactional data
Ans.Linkurious enterprise and the underlying graph database like Neo4j can be installed and configured on premises or could be managed on a private cloud. Alternatively if a customer is able to manage a hybrid cloud environment, where the cloud service provider guarantees the geographical bounds of the data centre, these tools will still work well.
The data will be analysed within your Linkurious instance. Linkurious does not store your data , it is stored in the native graph database like Neo4j. Linkurious only stores metadata. Data need not move outside the country at all, it depends how the storage infrastructure has been configured to store the data in the first place.
What volume of data have we compared to measure the performance?
Ans.Linkurious Enterprise is a visualization tool that depends on an underlying graph database like Neo4j.
Based on our experience of working with Neo4j graph database, we have customers whose graph databases have billions of nodes and relationships and their systems are performant. Large government enterprises, Banks and global enterprises are few such institutions who have implemented graph databases and use Linkurious enterprise to search and visualize patterns in the graph, the reporting performance is mainly the reason why they have shifted from a legacy sql/nosql database to graph database.
Can it be configured for Malwares?
Ans.If data can be prepared in Graph DB then the answer is yes.
Currently our customers are using Linkurious Enterprise in different domains like
Financial Institutions : Financial crime , AML
IT : Data governance , Cybersecurity
We would be able to help you better if you could elaborate on the use case you had in mind.
Where does this data get analysed ? In India or at your office in Paris? If so, data moves outside the country! In India, millions of transactions are taking place every day across various branches. Out of these , few lakhs of transactions are getting flagged as suspicious transactions. What is the capacity of Linkurious in terms of handling transactions.
Ans:Linkurious enterprise and the underlying graph database like Neo4j can be installed and configured on premises or could be managed on a private cloud. Alternatively if a customer is able to manage a hybrid cloud environment, where the cloud service provider guarantees the geographical bounds of the data centre, these tools will still work well. This data will be analysed within your Linkurious instance. Linkurious does not store your data , it is stored in the native graph database like Neo4j. Linkurious only stores metadata. Data need not move outside the country at all, it depends how the storage infrastructure has been configured to store the data in the first place .
We have active examples of both Linkurious and Neo4j handling billions of nodes and relationships and still delivering performant reports.
In terms of handling the transactions Linkurious has capabilities & features to handle large volumes. An interesting example is the investigation of Panama papers by ICIJ using Linkurious.
From a technical point of view, performance for the graph exploration (expanding nodes, executing queries, ...) depends on the underlying graph database: all the actions executed in Linkurious Enterprise are translated in queries for the graph database; Neo4j and CosmosDB provide very good performance even with very large graphs. Moreover, Linkurious Enterprise’ indexing strategy further speeds up some graph operations (such as a single node search). Regarding the performance of the graphical rendering of the graphs, we use Ogma (our proprietary library) the workload is distributed across users’ browsers. A benchmark is available here: https://doc.linkurio.us/ogma/latest/examples/performance.html
Can you briefly explain what happens when Linkurious is used to visualise data from Power BI? Does the data remain at source or is it transformed in a graph database and stored elsewhere? thank you
Ans: PowerBI acts as reporting tools: all the information displayed in the PowerBI dashboard are stored in (or exported from) an already existing database (relational or graph database). With Linkurious Enterprise is possible to use some values of the PowerBI’s dashboard as input for populating a new visualization or for re-opening a saved one: the data displayed in this visualization will always come from a graph database defined as datasource in Linkurious Enterprise
Just curious - looking for a specific customer that's good but will linkurious support monitoring a set of suspicious rings/people?
Ans: With Linkurious you can take several approaches to investigate suspicious customers.
-Either you can look for a particular customer using search and analyse in Linkurious interface to detect his connections and potential fraud rings.
- Other options could be Alerts in Linkurious. Alerts are a way to scan your graph database for specific patterns and display a match when such patterns appear in the data. A simple triage interface lets the users navigate through pattern matches and flag them as confirmed or dismissed.
Blog from one of our Customers: https://linkurio.us/blog/zurich-insurance-insurance-claims-fraud-prevention/