Confluent Reviews & Product Details

Confluent now offers MCP support for AI agents: an open-source local MCP server, a managed MCP server in Confluent Cloud, and Agent Skills. This lets AI coding assistants explore topics, schemas and connectors and help troubleshoot directly from the editor, which fits well with how we develop today. Combined with the fully managed Kafka service, we can spend even more time building applications instead of managing infrastructure. Review collected by and hosted on G2.com.

The managed MCP server is read-only, so for write operations you still need to run the open-source local server yourself. Beyond that, the earlier point still stands: a fully managed service creates a dependency on the provider, with some limitations and a cost. Review collected by and hosted on G2.com.

Smooth integration, and resilence aThe best part is that it removes the operational burden of running Kafka ourselves. We use Kafka as the backbone for subscription lifecycle events — membership state transitions, billing events, feature-specific event streams — and with Confluent we get elastic, fully managed clusters without dedicating engineering time to broker management, partition rebalancing, or upgrade cycles. For a lean distributed team, not having a Kafka ops rotation is a genuine capacity win.nd stability. Review collected by and hosted on G2.com.

Debugging tooling in the console is limited. Inspecting messages on a topic, tracing a specific event through the system, or answering "did this event actually get produced, and who consumed it" still means reaching for CLI tools or custom consumers. For a platform this mature, message browsing and search in the UI feel underpowered. Review collected by and hosted on G2.com.

With the real world change of requirements and moving to event streaming, the Confluent product can make a lot of useful with having streaming house framework and producers and consumers integration, it can integrate across the platforms more agile, seamless with zero loss of events, parallelism, and best retention. Review collected by and hosted on G2.com.

Basically the brokers and partitioning technology need to be improved a little bit as recovery sometimes gives issues. Even the partition leader with no more ZooKeeper improves a lot, but still there are issues that need to be fixed. Review collected by and hosted on G2.com.

I use Kafka’s pub-sub model when a file or process needs to be read and understood by multiple applications. I find Confluent easy to configure, and I don’t have to host or manage the server because others manage it for us, which makes it easy and convenient. We host it on the cloud, and Kafka is faster and better than IBM MQ. We process around 8 million documents a day, and Kafka is fast enough to keep up with that volume and process the files. I’m very happy with how the product works. Review collected by and hosted on G2.com.

Nothing Review collected by and hosted on G2.com.

I use Confluent to record our meetings and handle live-streaming events for the company. I like its ease of use and functionality. The source lets us publish live events in real time, and it’s powerful for our needs. It seems highly scalable, and its throughput and fault tolerance are impressive. Review collected by and hosted on G2.com.

I think the streaming process could be more multi-threaded. The functionality seems limited to a single stream instead of allowing multiple streams across multiple devices at one time. Review collected by and hosted on G2.com.

I use Confluent Kafka with my customers, and it’s the best choice for their strategy because it’s very easy to use and essentially plug-and-play. Review collected by and hosted on G2.com.

I dislike platform training because it doesn’t prepare my customers to solve real-world problems. Review collected by and hosted on G2.com.

I like how with Confluent, the entire setup can be done from the user interface or API calls, avoiding the headache of managing a self-managed cluster. It's really convenient because you just need to provision the Kafka topic, partition, and consumer groups, and it's ready to use. I appreciate the self-managed clusters, as I don't need to manage the backend infrastructure like other message queue services. The cross-cluster replication is a good feature, which is useful for disaster recovery and other activities. Additionally, estimating the volume of data and provisioning nodes accordingly is straightforward. You can spin it up in 1-2 days, and it's ready for all producers and consumers. Monitoring metrics from the UI and setting up alerts based on specific criteria is also a nice touch. Review collected by and hosted on G2.com.

For authentication, you cannot use Keytab-based authentication, which is commonly used. Instead, you have to use service account-based authentication, which is generally available. If you want to use Keytab-based authentication, you need to use Confluent Mesh, which is again a paid software. I think Confluent Mesh should be part of the Confluent package itself instead of buying that package separately. Review collected by and hosted on G2.com.

AI contextual stream processing is good for the latest trends and real-time recommendation engine Review collected by and hosted on G2.com.

The direct integration to the context-aware LLMs can be improved. Review collected by and hosted on G2.com.

It’s constantly being updated and improved, with new features and enhancements that make the platform better over time. Review collected by and hosted on G2.com.

More flexibility and the option to make 100% customizable connector from cloud Review collected by and hosted on G2.com.

I like the attention and services available from Confluent, especially those for critical services. It provides me with agility in services. The initial setup was easy because I received guidance. Review collected by and hosted on G2.com.

I would like Confluent to better specify the licensing issues. Review collected by and hosted on G2.com.