Visual graph analysis for Amazon Neptune
Kineviz (formerly GraphXR) connects to Neptune today. Query the graph, build a dashboard, bring in outside data, run the analysis and produce the report, from a desktop app you download or self-hosted inside your own VPC.
What Kineviz adds
A visual layer on the graph Neptune already runs
Neptune stores the relationships and answers fast at scale, which is the hard part. Kineviz takes what comes back and shows it to an analyst as a network they can rotate, filter and follow.
The whole network at once
Spread a result out in 2D or 3D, play it forward through time, lay it on a map. Filter to the ten entities that matter, then widen back out without losing your place.
Follow a lead, then write the query
Expand a node, see what it touches, keep pulling. Exploration is how an analyst works out which question is worth asking, and which traversal asks it.
Shareable with the whole team
Supervisors and partner teams open the same workspace in a browser. Reading the finding takes no query language, so the analysis reaches the people who act on it.
End to end
From a Neptune endpoint to a finished report
The whole workflow runs in one place, so an analyst never has to move the case between tools to finish it.
Connect to Neptune
Point Kineviz at your cluster and work against the graph in place.
Run your queries
Gremlin or openCypher against your Neptune endpoint, with the result drawn as a network.
Build a dashboard
Charts and views beside the graph, so the numbers and the network are read together.
Bring in outside data
Add files, APIs and third-party data alongside the Neptune graph, and merge the entities.
Do the analysis
Community detection, centrality and shortest path, with the Kineviz Agent running the steps and showing each one.
Generate the report
Hand the finding to the people who act on it, with every claim traceable back to the source data.
Two ways to run it
Start on a laptop, or deploy beside the cluster
Neptune customers arrive with very different constraints. Both routes read the same graph and do the same analysis.
Kineviz Desktop
Download it, point it at Neptune and start. An embedded graph database and a built-in agent come with it, so there is nothing to stand up first.
- Runs on your own machine, with no deployment project
- Connects to Neptune and analyzes in place
- In-app sandboxes and tutorials to get moving
- Good for a proof of concept or a one-off investigation
Kineviz Enterprise
Neptune keeps a cluster private unless you choose otherwise, which suits the data our customers work with. Enterprise is self-hosted, so it deploys into the same VPC as the cluster.
- Data stays in your VPC, behind the endpoint you already have
- SSO over SAML or OIDC, role-based and row-level access
- Audit logging of who looked at what
- Regulated, classified, sovereign and air-gapped environments
Use cases
The graph work Neptune customers already do
Fraud and financial crime
Rings, shared accounts, coordinated behavior and money flows that row-and-column tools miss.
Identity and customer 360
Resolving entities across sources, and following the journey that connects them.
Knowledge graphs and ontologies
Navigating a modeled domain, and building one from documents that were never structured.
Investigations
Case work that pulls outside data alongside the graph and has to stand up to scrutiny.
Across the Neptune family
Works with Amazon Neptune Database and Amazon Neptune Analytics
Kineviz connects to both service types directly, over Gremlin or openCypher depending on what the endpoint supports. The two are built for different jobs. The scale figures below are AWS’s own.
| Amazon Neptune Database | Amazon Neptune Analytics | |
|---|---|---|
| Built for | Mission-critical systems at large scale | Graph analytics, graph algorithms and vector search, in memory |
| Query languages | openCypher, Apache TinkerPop Gremlin, SPARQL 1.1 | openCypher |
| Access | Private by default. Every instance is created in a VPC and reached over HTTPS on the port set at creation. An optional public endpoint is off by default and requires IAM authentication | Private by default via a VPC interface endpoint, with an optional public endpoint. All requests SigV4-signed |
| Scale AWS publishes | Tens of billions of relationships; up to 128 TiB per cluster; up to 15 read replicas | Thousands of analytic queries per second |
| Where Kineviz connects | Gremlin or openCypher endpoint, from the desktop or from inside the VPC | openCypher over the graph endpoint |
Worth building on
Neptune Analytics indexes embeddings of up to 65,535 dimensions and queries them from openCypher, and Neptune offers fully managed GraphRAG with Amazon Bedrock Knowledge Bases. Kineviz builds knowledge maps from the same kinds of source: documents, emails and reports.
Sources: Amazon Neptune FAQs, the Amazon Neptune product page, and the AWS documentation on public endpoints and the vector index, read 10 November 2026.
Frequently Asked Questions
Does Kineviz work with Amazon Neptune?
Yes. Kineviz connects directly to both Amazon Neptune Database and Neptune Analytics, over Gremlin or openCypher depending on the endpoint, with results drawn as an interactive graph rather than returned as rows. You can run it from Kineviz Desktop on your own machine, or deploy Kineviz Enterprise inside the same VPC as the cluster. The full workflow runs in one workspace: connecting, querying, building a dashboard, bringing in outside data, running the analysis and generating a report.
How does Kineviz connect to Amazon Neptune securely?
By running inside the same boundary. Both services are private by default: AWS creates every Neptune Database instance in a VPC, reached over the HTTPS port set when the database is created, and Neptune Analytics uses a VPC interface endpoint with SigV4-signed requests. Kineviz Enterprise is self-hosted, so it deploys into that same VPC and queries Neptune directly. Your data stays inside the VPC, and analysts work without a public endpoint.
What is the difference between Neptune Database and Neptune Analytics?
Neptune Database is the fully managed graph database for mission-critical workloads; it supports openCypher, Apache TinkerPop Gremlin and SPARQL 1.1, scales to tens of billions of relationships and up to 128 TiB per cluster, and supports up to 15 read replicas. Neptune Analytics is an analytics engine for graph algorithms and vector search over data in Amazon S3 or a Neptune database; it is queried with openCypher and indexes embeddings of up to 65,535 dimensions. Kineviz connects to both.
Which query language does Kineviz use with Amazon Neptune?
Gremlin or openCypher. Kineviz checks what the endpoint supports and works with it. Neptune Database offers both, and Neptune Analytics is queried with openCypher. Teams keep working in the language they already write, and see the result as a network.
Can we try it without a deployment project?
Yes. Kineviz Desktop is a downloadable app that connects to Neptune from your own machine, with an embedded graph database and a built-in AI agent, plus in-app sandboxes and tutorials. It suits a proof of concept or a one-off investigation, with no infrastructure to stand up first.
Can we run this in a regulated or air-gapped environment?
Yes. Kineviz Enterprise installs on your own infrastructure, on-prem or in your private cloud, including fully air-gapped networks with no outside connectivity. You control who sees what through role-based and row-level access, sign-in runs through your SSO over SAML or OIDC, and every action is recorded in an audit log.
Bring a graph you already know
Pick a Neptune workload your team understands inside out. We will show you what it looks like in Kineviz.
Amazon Web Services, AWS, Amazon Neptune, Amazon Bedrock, Amazon S3 and related service names are trademarks of Amazon.com, Inc. or its affiliates. Kineviz is not affiliated with, endorsed by, or sponsored by Amazon.com, Inc.