Reviews from AWS Marketplace
0 AWS reviews
-
5 star0
-
4 star0
-
3 star0
-
2 star0
-
1 star0
External reviews
External reviews are not included in the AWS star rating for the product.
All connected
What do you like best about the product?
Easy and fast to deploy in productions. Many features to try and implement.
What do you dislike about the product?
Nothing in this moment. I only have used Db features
What problems is the product solving and how is that benefiting you?
Knowledge graph
- Leave a Comment |
- Mark review as helpful
Using neo4j in academic and commercial settings
What do you like best about the product?
I have tended to work with ontology data over the years. This data does not lend itself to RDBMS. Neo4j provides an excellent resource for storing the data, querying, and building applications on top of ontology data. I have used the community edition while working in academia with much success. There is just enough information on the web to get something up and running without direct sales support. I have moved to a commercial setting recently. We are exploring how to use neo4j to add an important element to our query engine.
What do you dislike about the product?
There seems to be a lot of pain involved in upgrading the neo4j system between versions (ex: 4.1->4.2->4.3). The upgrade path is brittle and very error prone. Simplifying this process is very important. I do not think the upgrade path should necessarily allow users to "skip" versions as part of the upgrade, but I think some of the steps could be automated.
What problems is the product solving and how is that benefiting you?
I want to build a graph of biological relationships (ex: genes, proteins, etc.) on top of flat file data from our clients. The goal is to augment their biological data with publicly available data to provide comprehensive insights into their data.
Good and fun!
What do you like best about the product?
Connection, relationship discovery, apoc, data science applications.
Profiler is a life saver.
Profiler is a life saver.
What do you dislike about the product?
Bloom doesn't let me create new node types / relationships unless they exist currently in the graph.
The neo4j dekstop / web app performs quite terribly when loading data above a certain limit. Even if it's not in the node visualization tab. Not sure what the threshold is, but it's fairly frustrating (I assume it's due to unoptomized js).
Option to return only what I queried for in the graph at first load... For example, I might query (n:Person)--(m:Movie) and get back the relationships in the visualization by default, which is generally great, but sometimes I want to quickly see if I'm missing something in the return statement. Although I understand that adding that feature may cause more confusion than it's worth!
The neo4j dekstop / web app performs quite terribly when loading data above a certain limit. Even if it's not in the node visualization tab. Not sure what the threshold is, but it's fairly frustrating (I assume it's due to unoptomized js).
Option to return only what I queried for in the graph at first load... For example, I might query (n:Person)--(m:Movie) and get back the relationships in the visualization by default, which is generally great, but sometimes I want to quickly see if I'm missing something in the return statement. Although I understand that adding that feature may cause more confusion than it's worth!
What problems is the product solving and how is that benefiting you?
Modeling IoT device data and the device's relationship to actual entites along with the operational context. The performance has been good and, honestly, allows me to get away with more than I probably should. But I won't complain about that.
Great on reads and modeling needs improvements writing at scale
What do you like best about the product?
So really like the property container. Modeling relationships between various types and the flexibility. Unlike SQL, Neo4j is able to do much more complicated join queries at scale.
What do you dislike about the product?
Write performance and locking are the most problematic. The pesimistic locking is very restrictive if that was configurable that would be a huge improvement. Also the clustering is much more the application's responsiblity.
What problems is the product solving and how is that benefiting you?
The problems that Neo4j helps the most with is modeling a dynamic application with dynamic reporting/dashboaring. It allows our customers to proactively add more properties and queries easily.
Great database, but there are still some issues
What do you like best about the product?
Neo4j has an opensoruce version. Support is good and the community around this database is pretty big. Cypher QL is easy to learn and use. Neo4j browser is user-friendly and has a lot of setting for both new and experienced customers.
What do you dislike about the product?
There are still some annoying bugs, like queries that can't be terminated by the kill command, in one of the recent versions I couldn't have more than 1 tabs open at the same time, etc.....
What problems is the product solving and how is that benefiting you?
I'm using Neo4j at work to store information about product catalog items and their relationships. We've got millions of nodes and hundreds of millions of relationships in our catalog, and Neo4j has proven its ability to easily handle it.
Best-In-Class Graph Database With a Gentle Learning Curve
What do you like best about the product?
I have been working with Neo4j professionally since 2020, primarily for the construction of knowledge graphs for regulatory reporting. Lots of things to like about the platform:
* Plenty of quickstart tools - particularly the Bloom browser, the desktop client, and Python tutorials make it easy to pick up as your first graph DB. It's a very accessible tool.
* Rich connector ecosystem for APIs, data sources, and data science capabilities.
* Speed of execution (vs. solving the same problem in a relational database).
* Plenty of quickstart tools - particularly the Bloom browser, the desktop client, and Python tutorials make it easy to pick up as your first graph DB. It's a very accessible tool.
* Rich connector ecosystem for APIs, data sources, and data science capabilities.
* Speed of execution (vs. solving the same problem in a relational database).
What do you dislike about the product?
Not many negatives here. The managed offering (Aura) could be a bit more backward-compatible - as it stands, it is VERY forward-looking and as a result, it does not support older Neo4j versions (==3.x) or the bolt:// protocol. It would be good to see Neo4j 3.x on Aura, as this is the version that a popular OSS data catalog tool (Amundsen) requires.
What problems is the product solving and how is that benefiting you?
Knowledge graphs - specifically, uncovering relationships between regulatory requirements/questions that are being asked by Environmental, Social, and Governance (ESG) regulatory bodies.
Thinking in graphs opens a whole new world
What do you like best about the product?
Coming from a traditionally RDBMS background, neo4j and graph databases have opened up new lines of thinking. The tools provided by neo4j make it very easy to get up and running quickly, and the free training programs they offer are second to none. They have well-documented drivers for a variety of popular languages.
What do you dislike about the product?
I'm not sure I can point to any particular things I dislike. It does require a shift in thinking from traditional RDBMS systems because you have to think more carefully about the problems you are trying to solve and the relationships between your data.
What problems is the product solving and how is that benefiting you?
neo4j is helping my team and me solve problems related to complex queries in a very large data set where how different data points are related to one another is key.
Supply Chain applications - initial evaluation
What do you like best about the product?
cypher is very easy to pick up and intuitive. browser portal is also easy to use. sandbox with examples and tutorials with example datasets is very useful for beginners.
What do you dislike about the product?
some complex queries are not easy to write and often reside in APOC library which for a new user "seems" intimidating.
What problems is the product solving and how is that benefiting you?
End to End supply chain modeling and supply chain event response system.
Ideally suited to solve previously hard problems
What do you like best about the product?
PRoblems that previously required using a graph library and had to be executed in memory are now much easier for us to solve in our Java application. These are in the domain areas of risk analysis and compliance.
What do you dislike about the product?
Some essential features missing from community edition. Aura DB free has a low memory ceiling of 250MB that makes it necessary to upgrade even for basic data loading fairly soon. While I understand that you need to incentivize people to upgrade to Enterprise and/or paid Aura tiers, these are limiting community adoption IMHO.
What problems is the product solving and how is that benefiting you?
Information security risk analysis: inheritance of potential impact levels and probability and risk calculation in a dependency graph of components. Mappings between different cybersecurity standards for compliance reports. Also saving audit trails in a temporal graph that allows what-if analysis and rollback to previous versions of relevant entities the transactional database.
Native Graphdatabase, scalable, performant, professional service and support
What do you like best about the product?
Open-source community edition helps to start with Neo4j. Easy install, many tutorials and community blogs. Enterprise edition comes with additional features and professional support.
The native graph database is flexible, scalable and can easily transferred in a cloud instance, if necessary. Good documentation, as well as a big community.
ACID-compliant database.
In addition to the database, a native integration of GraphDataScience algorithms for machine learning, graph embeddings.
The database comes with a visualization, making it easy to start and to provide users with a user interface
Professional services, boot camps and other workshops can be booked.
There are cloud versions available for Google Cloud and AmazonWebServices.
The native graph database is flexible, scalable and can easily transferred in a cloud instance, if necessary. Good documentation, as well as a big community.
ACID-compliant database.
In addition to the database, a native integration of GraphDataScience algorithms for machine learning, graph embeddings.
The database comes with a visualization, making it easy to start and to provide users with a user interface
Professional services, boot camps and other workshops can be booked.
There are cloud versions available for Google Cloud and AmazonWebServices.
What do you dislike about the product?
Some of the features are only available for enterprise version including live-backup and role model.
While live-backup can be performed on a running database, for community edition the database has to be stopped. Which is ok for pilots and smaller systems. If you can not allow down time of the database you have to go for enterprise.
The role model can be handeled by an external or additional application layer.
While live-backup can be performed on a running database, for community edition the database has to be stopped. Which is ok for pilots and smaller systems. If you can not allow down time of the database you have to go for enterprise.
The role model can be handeled by an external or additional application layer.
What problems is the product solving and how is that benefiting you?
First of all the most important - connecting data (breaking data silos). Starting with meta data management which is obviously connected to raw data.
Connecting public data with in-house data we created a knowledge graph.
Connecting sensitive data from patients with public knowledge from literature.
Machine Learning on public and in-house data - including clustering, classification.
Connecting data from animal models, ontologies with patient data for translational research
Connecting public data with in-house data we created a knowledge graph.
Connecting sensitive data from patients with public knowledge from literature.
Machine Learning on public and in-house data - including clustering, classification.
Connecting data from animal models, ontologies with patient data for translational research
Recommendations to others considering the product:
Start with the community edition and built a quick pilot with your users. Involve them in data modeling - which is in most cases 99% of what they draw on white boards. Evolve from there.
showing 61 - 70