Share your experience using Plutora

My main use case for LaunchDarkly is managing the controlled rollout of capability in my in-house application named Escrow Navigator, a conversational assistant that helps the escrow team retrieve and interpret order information and documents. My use case includes enabling new document upload capability, introducing support for additional file formats, and allowing users to ask questions about newly uploaded documents without selecting them again. As a Lead Business Analyst, I define rollout requirements, document acceptance criteria, and coordinate validation with development, QA, and business stakeholders, focusing on feature behavior and release readiness rather than SDK implementation.

LaunchDarkly helps us manage the rollout of new features by using feature flags and user tagging, which are the most valuable capabilities for our workflow. These capabilities allow us to make a capability available to a pilot group while other users continue with the existing experience. For example, when introducing questions about newly uploaded files, we can enable the capability for designated testers, allowing them to validate upload status, document processing, conversation association, and citation before we expand access. This gives us a practical way to gather feedback on a complete business workflow.

LaunchDarkly has positively impacted our organization by helping our team introduce new capabilities through controlled phase releases, making features available to selected users, gathering feedback, and validating business requirements before expanding access. This improves coordination between business development and QA teams and gives us greater control over release readiness. The main organizational benefit is increased confidence when introducing changes while limiting the number of users exposed to potential issues.

We are currently tracking logs in Power BI to see how much feedback we are receiving, and negative feedback has recently been decreasing. We are more focused on UAT, allowing selected testers to validate newly uploaded questions, citation accuracy, and process status against defined acceptance criteria, which contributes to a greater impact on ROI in our project.

We have seen great improvement in metrics such as token input and output, and latency has shown significant improvement, particularly in the production environment.

The best features LaunchDarkly offers include giving us more flexibility in controlling technical deployment with business readiness. A feature can be deployed while access remains limited to the user responsible for validating it. For Escrow Navigator, this supports phase validation for document upload and processing changes, allowing business users to review the experience while QA checks acceptance criteria before broader availability. The main benefit is better control over feature exposure and release coordination, although we have not formally qualified the financial benefit.

The flexibility helps our team when rolling out features by enabling us to roll out a feature to a lower environment and disable it in a higher environment after some time, allowing us to manage features without conducting a major deployment.

Regarding the feature flag, in the UI, if we have enabled a feature such as giving feedback, but we want to hide that feature, using LaunchDarkly, we can disable that flag, and it will reflect in the UI immediately. We also manage document upload and document review features via flags, with LaunchDarkly helping us by providing more flexibility in coordinating technical deployment.

I use LaunchDarkly with GitHub. Our development team utilizes LaunchDarkly alongside GitHub, where GitHub manages our project source code and pull requests while LaunchDarkly controls the availability of features implemented behind flags. This allows teams to merge and deploy code, enabling capability for selected testers before a wider rollout. My role focuses on requirements and UAT validation while the development team handles technical integration.

From a business validation perspective, I would find a consolidated comparison of flag settings across environments helpful. Understanding differences in configuration and targeting when reviewing a feature in dev, UAT, and production is important, along with clear business descriptions, ownership information, and visibility into flags that are ready for retirement. This would make release review easier for stakeholders who do not work with technical configurations every day.

If there are links to documentation available, that would be really helpful, especially from a knowledge transfer perspective.

I have been using LaunchDarkly for around one to two years.

LaunchDarkly is currently stable, and we have not faced any issues with reliability or downtime.

From our project perspective, LaunchDarkly has supported our feature management across environments and pilot user groups without accessibility issues. As usage grows, maintaining clear flag ownership, naming conventions, and cleanup practices becomes essential. I have not personally conducted load testing, and my assessment is based on business validation and release activities.

We have not faced issues with customer support, as we have raised tickets and received solutions within the expected SLA timeframe. I also coordinate internally with the development and platform teams who handle any vendor escalations, leading to a good experience overall.

I rate customer support an eight out of ten.

Prior to LaunchDarkly, we used an in-house tool that managed feature availability through application configuration settings and coordinated deployments. This led to difficulties as enabling a feature for selected pilot users required additional deployment or configuration work. LaunchDarkly gives us a centralized way to manage feature availability and targeting, making phased rollout much easier to coordinate.

We previously used an in-house tool for delegated feature management that managed feature availability. My assessment of LaunchDarkly focuses on its support for our requirements in controlled rollout, pilot testing, and business validation, as we no longer need multiple deployments across environments for managing feature access. LaunchDarkly enables us to disable features for users from its settings without the need for deployments.

We definitely see a return on investment, with operational values in easier pilot releases, more focused UAT, and greater control over feature availability, although we have not formally measured financial ROI or confirmed specific amounts saved. The benefits assessed could reduce release coordination efforts and rework, but supporting data is needed to translate them into concrete cost savings.

From a BA perspective, I assess value through the flexibility it provides for controlling release and pilot testing. My role in setup revolves around defining rollout requirements, identifying pilot users, and validating the enabling and disabling of feature behavior in Escrow Navigator, as the development team manages technical integration and configuration.

Compared to our previous approach of managing features through application configuration and deployment, LaunchDarkly stands out by offering more flexibility for controlled release and pilot testing. For Escrow Navigator, its main value is enabling selected users to validate new capabilities before wider availability, fitting projects that need phased rollout and clear release governance, though I have not conducted a formal evaluation of competing platforms.

I would rate LaunchDarkly an eight out of ten.

I chose an eight because of the flexibility I have discussed earlier, along with the capability of supporting multiple environments, which I find good. However, since I am not sure about pricing, I do not give it a ten. In the future, when I have more understanding of pricing, I would definitely want to explore it and potentially change my rating.

In terms of accuracy and reliability, I find LaunchDarkly's capabilities to be really good, as it supports governance and indirectly enables our application changes for certain testers to validate extra data and documents. While it would not directly improve dependence on document extraction or validation, any measured accuracy improvement that is considered ground truth is very helpful for us.

Advice for others looking into using LaunchDarkly includes starting with a small pilot and a clear use case, such as enabling a new capability for selected users before a wider release. I recommend defining flag ownership, naming conventions, and cleanup responsibilities early, including both enabled and disabled behaviors in acceptance criteria during UAT, validating fallback behaviors, while keeping feature targeting separate from application security permissions. Finally, review pricing against expected user measures to assess value.

My main use case for LaunchDarkly is feature gating. Whenever I introduce a new feature, I put it behind a LaunchDarkly flag and then use it to release that feature percentage-wise. I release it for some percentage of people or some team members, then see if everything works fine, and then roll it out for 100% of the users. I also use it for A/B testing.

A quick specific example of a feature I released using LaunchDarkly is the new cancellation flow, which was basically to reduce some churn. I didn't roll it out for all users; I first tested with only basic plan users. When I saw that it was working fine for them, I rolled it out for another plan, like the professional plan, and so on. Then I introduced it for the enterprise plan users, and then it was rolled out for all the other plan users.

With AI coding tools helping my team ship code faster than ever, LaunchDarkly helps me maintain control once that code hits production by following a percentage-wise rollout strategy. I first roll out changes to a small percentage of users, then monitor them, and if everything works fine for them, I gradually increase the percentage.

With LaunchDarkly, I use Sentry for error logging and monitoring whatever goes on with the changes, and it helps in observability and catching and figuring out the errors or any issues the users are facing.

LaunchDarkly has positively impacted my organization as I have been able to test things to a small percentage of users before rolling it out to a larger audience. If I have a bug in my production rollout, the blast radius is very small with this percentage rollout strategy, and if I find something, I can just quickly roll back that.

From this approach, I can say that I have much less production incidents. In my four or five years here at my current company, I haven't seen any production issue where I have to roll out every change for all users because I catch it at an early stage as well because of LaunchDarkly, and that helps me improve my reliability metrics.

One of the best features that LaunchDarkly offers is that if I create a flag in one environment, such as beta, it is automatically created in other environments, in staging and production, and all I have to do is enable it in those environments whenever I want. Another feature I really appreciate is the simplicity of adding more conditions and making the flag more complex according to my use case. The interface is very straightforward, easy to understand, and intuitive.

The automatic flag creation across environments helps my workflow because, as I mentioned before, if I create a flag and test some rollout strategy, I would like to test it on my beta environment first, then on stage, and then finally roll it out to the production environment. If this feature was not there, I would have to create this flag three times in a row. However, since LaunchDarkly provides this feature, I create the flag only once, and then it is replicated to all the other environments, and then I can just turn it on in the other environments without creating it from scratch. The UI is seamless to take care of all of these things.

One improvement that I think LaunchDarkly could make is that the search can be improved. Sometimes I find it complex to search for my feature flags or the feature flags created by other users.

I have been using LaunchDarkly for three years.

LaunchDarkly is stable.

LaunchDarkly's scalability is good because I have a large user base of approximately 40 million users, and it scales pretty reliably.

Since my start at this company in 2022, my organization has been using LaunchDarkly only, so I don't have any context about what my company was using before LaunchDarkly. My organization didn't use any different solution before LaunchDarkly.

I can comment that the reliability improved a lot with using LaunchDarkly.

My organization purchased LaunchDarkly and received the license.

My organization directly chose it without evaluating other options.

On a scale of 1 to 10, I would rate LaunchDarkly a nine. I picked nine out of 10 because of the search feature I mentioned.

LaunchDarkly is deployed in my organization as a public private cloud.

I advise others looking into using LaunchDarkly to increase the reliability of their product by adopting a percentage-wise rollout strategy. You can also use A/B testing to figure out which features are liked more by the users, so it's a very good product to do all of these things.