The biggest win has been operational: Redpanda is a single binary with no JVM and no ZooKeeper/KRaft quorum to manage separately. Our previous Kafka cluster needed brokers plus ZooKeeper nodes.

Integrations: Kafka API compatibility was the deciding factor.

UI/UX: Redpanda Console gets used daily to check consumer lag, inspect messages on a topic, and confirm schema registry entries without dropping to rpk. Being able to hand a non-engineer a URL to see whether events are flowing has cut a real amount of Slack back-and-forth.

Performance: p99 produce latency sits sustained. Predictability matters more to us than peak throughput.

Pricing / ROI: Fewer, smaller nodes for the same workload dropped our streaming infra spend.

Support / Onboarding: Always respond fast, shared a Slack channel with high quality of the migration help.

AI/Intelligence: Haven't used too much but will definitely try something out. Review collected by and hosted on G2.com.

Observability: Metrics are Prometheus-native and thorough, but mapping them onto the Kafka dashboards and alerts we already had took real effort , names and semantics don't line up one-to-one. Review collected by and hosted on G2.com.

What I like best about Redpanda Streaming is that it gives us a Kafka-compatible platform with strong performance, better operational simplicity, and a clear positive impact on infrastructure costs. In our case, this was very valuable because we could migrate existing Kafka-based services without fully redesigning the way teams produce and consume messages. Redpanda has helped us modernize our streaming platform and reduce part of the complexity and cost related to our previous Kafka architecture, especially around multi-region traffic and infrastructure operations. The BYOC model also fits well with our infrastructure strategy, as we can run the platform close to our workloads while still getting vendor support and managed operational guidance. Another important point is the performance and stability improvements we have seen during the migration, especially after tuning configuration and working with the Redpanda team on specific issues. The onboarding and support experience has also been useful: when we found problems around retention, disk usage, or producer behavior, the Redpanda team helped us understand the root cause and apply the right changes. Overall, Redpanda has helped us improve our real-time streaming platform while reducing operational burden and infrastructure costs compared with our previous Kafka setup. Review collected by and hosted on G2.com.

The main thing I dislike about Redpanda Streaming is that some operational behaviors are still not always easy to understand or predict without help from the Redpanda team. During our migration, we found some cases where disk usage, local retention, and cleanup behavior were not as clear as expected. For example, we had topics configured with local retention, but older segments were still kept on disk until additional settings were adjusted. This created some uncertainty during production operations because it was not obvious from the beginning why disk usage was not decreasing. Another area that could be improved is the Terraform provider and infrastructure-as-code experience. Since we manage most of our infrastructure through Terraform, provider stability, complete resource coverage, clear examples, and predictable behavior are very important for us. Some parts of the provisioning and configuration flow still feel less mature than the core streaming product, which can slow down platform teams when automating environments or applying changes safely. I think the product would benefit from clearer documentation, better defaults, stronger Terraform provider support, and more visible explanations in the UI around retention, tiered storage, housekeeping, and space management decisions. The support team has been helpful, but ideally the platform should make it easier for users to understand and operate these situations directly through metrics, alerts, recommendations, and reliable automation tooling. Review collected by and hosted on G2.com.

UI / UX

Redpanda Console is clean and intuitive. I can browse topics, check consumer group lag, and get basic cluster configuration at a glance without digging through config files. For day-to-day monitoring this saves me from running repeated rpk commands just to sanity-check the cluster state. Deeper operations like ACL management still push me to the CLI, but for the routine checks I do multiple times a day, the Console cuts that time significantly.

Integrations

The Kafka API compatibility means we can integrate Redpanda with many tools/technologies such as Kroxylicious (a Kafka protocol proxy) without any changes to client code. That gave us tenant isolation and header enrichment at the proxy layer, something that would have taken weeks of refactoring to achieve natively in Kafka. Existing connectors, monitoring agents, and tooling all just worked from day one.

Performance

Broker restarts are noticeably fast compared to what I was used to with Kafka. Predicting workload is much more comfortable because the latency profile is consistent. I'm not constantly second-guessing whether a spike is GC or actual load.

Pricing / ROI

Best example is the Community edition covers most production use cases at zero licensing cost, which I've seen benefit multiple organisations I've worked at. It removes the conversation about budget approval before you can even prove the technology works.

Support / Onboarding

Documentation is well-structured and developer-friendly. I can usually find what I need without raising a ticket. Moreover, when I needed to raise a Redpanda Support Ticket, I asked its built-in AI and it gave me needed solution. The Redpanda community Slack is active and I've had blockers resolved same day. When we do need enterprise-level help, our Principal Customer Success Manager is genuinely engaged with our use case, not just routing us to generic docs.

AI / Intelligence

The built-in AI for support is already useful, but there's a next step that I'd love to see: AI that can correlate metrics directly inside the Console and explain anomalies in plain language during an incident. Right now I'm context-switching between dashboards, logs, and external tools to piece together what's happening. Review collected by and hosted on G2.com.

The Helm chart configuration surface is large, and when you’re deploying across multiple environments and regions, it’s very easy to end up with a sprawl of values files that are difficult to diff and audit.

A more structured knowledge base for common operational problems would help a lot. It would cut down the time I spend asking questions that have almost certainly already been answered. Review collected by and hosted on G2.com.

That there are a lot of disk integrations that help with writes making things like disaster recover, IceBerg ingestion, very very fast, simple and their support teams helping us do that.

Most Kafka varients have connector platforms and those are helpful. But performant reliable disk level integrations I think postitions RedPanda at a better place. Review collected by and hosted on G2.com.

Some of the documentation for various layers of APIs are a little confusing. Similar to the connector platform documentation. People feel like they can administrate things via the UI. But sometimes I think it confuses the audience while giving them a lot of power. Review collected by and hosted on G2.com.

What I like best about Redpanda Streaming is how easy it is to deploy and manage, especially with the BYOC option. The onboarding experience is straightforward, the documentation is clear and practical, and the Terraform support makes it easy to automate deployments and manage the infrastructure as code.

It also integrates well with our existing monitoring solution, which made adoption much easier and allows us to keep using the observability tools and processes we already have in place.

The support experience has also been excellent. Having direct access to the team through Slack is very helpful, and responses are consistently quick and useful when we have questions or need assistance.

Overall, Redpanda significantly reduces the operational overhead of running a streaming platform while still giving us the control and flexibility we need in our own cloud environment. Review collected by and hosted on G2.com.

I don’t have any major complaints so far. One area that could become more important for us as we grow is having more self-service options for modifying existing clusters.

For example, being able to scale cluster size up or down, make certain networking changes, or adjust other infrastructure settings directly would be useful. To my knowledge, some of these changes currently require opening a support ticket rather than being handled through self-service.

The support team is responsive, so this hasn’t been a significant issue for us, but having more of these capabilities available directly would make managing the platform even easier as our usage scales. Review collected by and hosted on G2.com.

Pricing is very competitive, and the performance compared to other streaming technologies is at a very good level. The UI is intuitive and easy to navigate. Support is very fast and, most of the time, provides a clear way to find a solution. Review collected by and hosted on G2.com.

Sometimes it’s hard to figure out why something doesn’t work. Configuration errors should be clearly visible in the GUI. In our case, a single missing curly bracket meant Redpanda Connect wasn’t consuming from topics, but there was no error message that helped us identify the issue quickly. This suggestion doesn’t apply only to this specific area—errors and misconfigurations in general should be much more visible for customers. Review collected by and hosted on G2.com.

As a leading company in the internet gaming industry operating at a massive scale, we migrated our Kafka infrastructure from a major, well-known provider to RedPanda two years ago.

What we like best can be broken down into three key areas:

Pricing and Flexibility: The two primary drivers behind our migration were a more competitive pricing model and superior setup flexibility. The Bring Your Own Cloud (BYOC) model perfectly fits our needs—we retain control and pay for nodes and support, while they seamlessly maintain the software.

Flawless Onboarding: The initial setup and the entire migration process from our previous vendor went smoothly.

Exceptional Support: Throughout our 2-year partnership, their support team has been highly responsive and deeply committed. When operating at our scale, issues are inevitable, but RedPanda treats our challenges with the utmost seriousness and works shoulder-to-shoulder with our engineers to resolve them. Review collected by and hosted on G2.com.

Operating at our scale means technical challenges are inevitable. Recently, our main concern has been the quality of regular, scheduled migrations, which impacted our system's performance.

However, how Redpanda handled this downside reflects the strength of our partnership. They took our feedback very seriously—freezing the production cluster, investigating the root cause, and working shoulder-to-shoulder with our engineers. They successfully fixed the issues on the broker side and provided useful optimization suggestions for our consumer applications.

While the stability and impact of scheduled migrations at high scale is an area that requires continuous refinement, their proactive, transparent, and collaborative approach to solving the problem mitigated the issue effectively. Review collected by and hosted on G2.com.

I like Redpanda Streaming for its reliable fast streaming capabilities with large volumes. It provides strong performance and reliable event streaming capabilities, which is important for our real-time data processes. The initial setup was relatively straightforward for basic use cases. Review collected by and hosted on G2.com.

I find that Redpanda Streaming delivers strong performance and reliability, but I believe improvements in documentation, advanced integrations, monitoring capabilities, and ease of administration would help create a smoother user experience, especially for complex deployments. The platform requires a learning curve for new users, and advanced deployment, monitoring, and optimization may need additional expertise. Some integrations and tooling options could also be expanded further. Review collected by and hosted on G2.com.

What I like best about Redpanda Streaming is its operational simplicity combined with strong performance. It provides Kafka API compatibility while avoiding the complexity of managing ZooKeeper and reducing overall infrastructure overhead. The platform delivers predictable low-latency performance, scales well, and offers a straightforward developer experience. I also appreciate that it enables us to standardize on a single streaming platform for a wide range of use cases, from event-driven architectures to data integration, while maintaining full control over deployment and upgrades. Review collected by and hosted on G2.com.

What I dislike about Redpanda Streaming is that, while the core platform is mature and reliable, the surrounding ecosystem is not as extensive as Apache Kafka's. Some integrations, community resources, and third-party tooling are less common, which can occasionally require additional effort during implementation or troubleshooting. There are also areas where operational visibility and management capabilities could be further improved, particularly for large-scale enterprise deployments. Overall, these are relatively minor concerns compared to the benefits the platform provides. Review collected by and hosted on G2.com.

I really like Redpanda Streaming because it offers small latency, which enhances data accessibility and custom connectors. The event-driven architecture is a big plus, as it's Kafka API compatible without requiring Zookeeper, making it more lightweight. It also comes with a nice UI that allows me to filter data using JavaScript, which is better than what I've used with Confluent. Additionally, Redpanda Streaming's command-line interface, rpk, helps with certain low-level tasks. Their support is top-notch and readily available, answering all my questions no matter how big or small, and they also provide engineering expertise when needed. I appreciate their documentation and support team, which made the initial setup process manageable. Review collected by and hosted on G2.com.

Redpanda needs a data lineage graph that shows the data pipelines from ingress, to Kafka processors, producers and consumers, and egresses via connectors (to databases, etc). Also, sometimes the UI is a little 'hokie' when trying to find certain messages if there is a lot of traffic. Review collected by and hosted on G2.com.