CloudDat Reviews

At Flash AI, we are using CloudDat to build, fine-tune, and deploy multi-modal AI workloads and frameworks into our inference pipeline. Our compute requirements are extremely spiky, with instances where we need significant resources and other times when we do not. The distributed model training and data pre-processing clusters help us run intensively for four to twelve hours, followed by idle or lower workloads. We have implemented CloudDat on an hourly consumption model so that we have scalable data orchestration which is managed, with object ingestion included. We basically use it without doing a monthly fixed infrastructure allocation, which is our main use case.

The challenges are primarily due to the volumes coming in. Since we operate on an hourly metering system, we are seeing trends and optimizing per-hour metering. When there is a need, we scale up. When there is not a need and the volumes are low or the workload is low, we opt for less load or no load. This is how we manage that.

We have definitely seen things that have helped our organization positively. Machine learning engineers no longer need to wait for shared cluster queries. We can spin up parallel data clusters simultaneously on an hourly basis and shut them down as soon as the loss metrics coverage is met. There is true elasticity for our teams and engineers. We have also seen that CloudDat has reduced the idle resource waste in standard dedicated environments. Previously, we paid for oversight and idle time. The transition to hourly provisioning eliminated sixty percent of non-working hours' compute expenses. CloudDat has given us reliable API performance under load and very fast provisioning time, under sixty seconds for standard data clusters. It has ensured our automated workflows do not have any bottlenecks and developer momentum is maintained.

With idle resource waste, we measured it with both cost and timelines. As we transitioned to the hourly provisioning model, we only pay for the spikes and the things where the workload is, and we do not pay for anything unnecessary. This has given us a sixty percent saving on money for non-working hours where the data clusters were not working as per our expectations or they were running for an unnecessary reason, which was infrastructure overhead. This is how CloudDat has helped us.

Since engineers no longer have to wait for cluster queries, they can work faster and the bottlenecks have been resolved. Team collaboration has improved. Team morale has been elevated due to this.

The first and most important feature is the per-hour metering, the ability to provision high-throughput compute and storage volume. It is helping us to execute pipeline tasks and clear them down immediately with per-second rounding, billed strictly on an hourly scale. That is my first feature, which is my personal favorite. The second feature is native CI/CD and infrastructure as code. It is a first-class Terraform and OpenTofu provider support. We have seen it coupled with comprehensive CLI tooling. It is allowing our DevOps pipelines to dynamically trigger and destroy CloudDat clusters directly from GitHub actions without doing any manual work. The third feature, which is my personal favorite, is the active training dataset on high IOPS tiers. During compute runs, CloudDat automatically transitions to low-cost and archival states the moment instances detach, which is preventing phantom spend.

CloudDat has helped us with low code and infrastructure as code. The native CI/CD has helped our DevOps pipeline to trigger and destroy CloudDat clusters directly from GitHub actions. Basically, when there is a need for the cluster or a pod to be up, it is automatically triggered from the GitHub action. It is an automated tool solving the purpose within the software and following the native CI/CD pipeline approach, which is a great thing. Whenever there is a requirement, it is scaling up or scaling down simultaneously.

The governance and security are pretty strong. CloudDat provides AI infrastructure without slowing down machine learning delivery, which has given our DevOps team the perfect balance of elasticity, automation, and cost accountability in our organization.

The accuracy and reliability is one hundred percent. I feel that there is no challenge with using CloudDat's AI.

The integration with other AWS services is definitely seamless. It has given us exceptional integration workflows and connectors through CloudDat, which is doing really great.

CloudDat is a great solution and has given us a great return. One thing as a suggestion is that while tracking the hourly usage is very precise, the dashboard threshold alert can sometimes experience a lag of ten to fifteen minutes. This occasionally makes the proactive cut-off automation tighter than ideal. For example, if I am opting for a two-hour auto cut-off, sometimes it is detecting that it is there, and there is a fifteen-minute lag. The cut-off is basically either fifteen minutes before or fifteen minutes after. There is a chance of a mismatch in the billing due to this, which can be improved. The rest is fine.

The billing issue which I highlighted is something very critical sometimes. When we are a regular customer and we are working with CloudDat, sometimes it is very disappointing to see that there is a lag in the subscription. This is why I have given it a four.

We have been using CloudDat for the last one year.

CloudDat is one hundred percent stable.

CloudDat can easily scale up and down as per the requirements and needs. Scalability is a ten out of ten for me.

Customer support is really great. We had a discussion with them last month and they solved our query within forty-five minutes. That was a great experience. I would rate the customer support ten out of ten.

We did not use another solution previously.

Deployments were quite easy. We got it configured within two days. Since CloudDat is an hourly-based product, that was not at all difficult. Within two days everything was done.

It was easy to configure in our systems and organization. We did not feel any challenge or run into any problem.

We have seen a cost saving in our organization by switching to CloudDat's hourly-based billing model. It has reduced our infrastructure spend by seventy-two percent as compared to our previous baseline reserved instance approach. That was the first impact. Second, our engineering velocity has increased by seventy percent because the pipeline cycle times have roughly dropped by eighty to ninety percent, and the engineers can trigger full-scale pipelines and test pipelines on demand, rather than throttling the jobs through shared staging pools. Third is the payback period. We have seen the complete ROI recovered in the first forty-five days of the deployment, and it is driven primarily by the elimination of fortnight idle data of the compute instances.

Metering and billing experience is one of the best things we have seen there. Since we are going ahead with the hourly consumption model, it is pretty straightforward and saving a lot of time and money for our organization.

Pricing, setup cost, and licensing is very transparent since CloudDat is offering an hourly model with no hidden penalties or surcharges between the compute nodes. It is great to have this kind of model, avoiding multi-year contracts up front, which is saving a lot of time and money for us.

Before using CloudDat, we definitely compared other solutions such as Databricks, Snowflake, RunPod, and AWS EKS.

We found the procurement easy for the organization and did not run into a challenge while procuring CloudDat for the organization.

Anyone who is looking for an exceptional service without doing any commitment to the company should definitely see CloudDat's hourly model. That will be perfect for you. I would rate this product eight out of ten.

My main use case for CloudDat is storing, processing, and analyzing data. At my project and company, we use platforms like AWS instead of some servers. This really helps when data is handled in real-time. CloudDat's real power comes into picture in real-time scenarios where anyone can use cloud in a very safe way. Cloud is basically referred to as a more secure method. That's where in the real-time scenario, we use it.

I work in healthcare as of now, so my scenario with CloudDat is more into wearable devices. For example, for any patient, they have to check heart rate or any oxygen level. We have scenarios where we have wearable devices to stream patients' vitals. For that, we use these data flows in a cloud dashboard where doctors generally use this data flow, and that really benefits the health risk of that particular patient.

As of now in the health industry, this is one of the examples where I really come into the picture, and I have hands-on experience with CloudDat. I don't think I have any other good example other than this.

The best features CloudDat offers are very good. In real-time, it gives us a cloud platform where the data pipeline for real analytics is used. Some features I used include ones that helped me in healthcare IoT, and also examples into scalability, where anyone can handle this data in a cloud, and even if it doesn't go down, there won't be any downtime. Since it is a cloud platform and it is very fast for deploying any new feature, that is one of the best things which I have come into hands-on.

Using CloudDat has positively impacted my organization with more scalability for a long term. Handling became very easy for each user, and any kind of data, any large amount of data can be handled with different users in no time. I have come across a few things including more flexibility. Flexibility in the sense it supports both short-term and as well as long-term growth plans, where if you want to integrate any data through cloud, that can also happen. That's one of the good features which we experienced.

Improvement-wise for CloudDat, I think some days there is some latency. The scaling must be fast enough, but sometimes it avoids delays. That is one challenge we had. Sometimes the server might be down and there would be some downtime. That is one thing which we had a challenge with. When you use any policies, particular policies in that features, there would be some predictive analytics to improve that efficiency. It's more of having to always see what are the new features they have added and get advanced to that. That's one of the challenges we faced. In a real-time, one more thing is our project handles HIPAA, so it is always used by many users. Their scaling is also one of the features which we had a challenge with, actually.

Regarding needed improvements for CloudDat, documentation-wise, from cloud, there are certain features where you can directly download, but that would not help you actually. You have to always be updated to cloud features. There are certain versions every time it gets updated and you have to always get updated to those new features. Documentation matters from previous to new features, all that. In real-time, I maintain a documentation which ensures every time what was in past and what is present and what we can do more in the coming days so that it can be more clear so that we have anything handled in a performance way. It should be documented and it is already documented.

An improvement needed for CloudDat that I haven't mentioned yet is that it should secure more cloud dashboards to monitor patient vitals in real-time. As of now in IT, since I'm into healthcare, it generally helps with leaking risk of any particular patient's information, but in the dashboard, you will come to know everything. That's where I think it should hide. That's one feature I see which needs to be changed.

I have used CloudDat for more than two years, specifically for about two years.

CloudDat is stable.

CloudDat's scalability is very good because as I already told, it generally handles the data in the cloud. It handles data wisely. You don't need to worry more on how it is handled in a deeper way because it's all handled in a server. In a real-time, we basically handle the scalability in an upscale, where you don't need to upgrade your VMs always. It is always stored in an upscale. There are some other diagonal scaling as well, but we don't use them. There is also horizontal scaling as well that we don't come into picture with. In my company, presently we use vertical scaling and all that.

When I say scalability with CloudDat, it basically helped me handle millions of data without downtime, where I handle a lot of data. Whenever I create any data in the cloud, that is stored in that particular server. At a time, any other user can access that data which I have stored in the cloud. The scalability is very good with CloudDat. It doesn't shrink any kind of data and it ensures the performance and the data is more accurate and more efficient. That's where a real-time picture comes into picture. It's very long, it's very planned, it's very expected data.

Our customer support for CloudDat is okay. There is a support which generally supports based on the ticket raised and they generally come and schedule a call for what is more detail of asking what the requirement or what is the issue. Based on that, the support is always given. That's really good that the support is there.

I did not previously use a different solution before CloudDat, but my company seniors earlier were using a different project, a different cloud rather than AWS. So there was Azure and also there was Microsoft. They have used certain other features and then at last, they brought in CloudDat. Earlier it was different.

I think I have seen a return on investment with CloudDat because it really helped save a lot of time. The cloud platform enables different users. That's very useful and the data is always stored in the cloud. That's more of time-saving.

Before choosing CloudDat, I evaluated other options such as Microsoft Azure and Google Cloud Platform. These were two they had chosen earlier, but again, they thought that CloudDat is the best one. They came in a real-time to choose CloudDat.

I can share specific outcomes related to time saved with CloudDat. You don't need to always access it each time. If a user can have multiple access to a particular cloud, then he can run in multiple projects and that really helped each candidate or each person to access a cloud and save their time. Time-saving is one of the most crucial and one of the best features over here in CloudDat.

I would definitely say if you are using CloudDat where you need everyday data in a cloud and you want to use something, especially storage level where you have a lot of data, unlike other industries, especially when you are into healthcare and all that, instead of buying new servers, CloudDat helps you a lot. That's what my experience is. Regarding CloudDat's AI capabilities, I think its governance and security is very good, actually. Cloud has enabled those features in the platform where you don't need to worry about the governance. They always give a stable data where it's always protected and it's sensitive data, especially in my healthcare, it's a sensitive data. Whenever you download anything, it is not passed to anyone. It can be only downloaded from a particular user and it can be used in a real-time for that particular person. Governance and security-wise, it's very true. There won't be any leakage of any data, especially for the sensitive elements. It's very reliable towards what we need in the market.

Regarding CloudDat's AI capabilities, accuracy is also very good here. AI things should come more into picture in CloudDat. As of now, it handles a few things including continuous integration, it helps through the cloud. Some other features it handles include always maintaining some policies where you have to always continuously follow that. Cloud settings also plays a different role over here. These are more points which I have in a real-time.

The features of CloudDat, basically, since it's cloud, it is very secure. It is also more reliable where it has handled a lot of encryption of data. It identifies who is accessing the particular data in the cloud, including entity management, all that. Sensitive information would be there at my healthcare industry. We always give access to limited users. The scalability is quite important as well as it's more reliable. Data protection is more important at my present company and which the cloud handles. Some compliance and governance also it handles, where it is more accountable towards who handles that particular data and what kind of regulatory it handles. We handle more HIPAA files, so that is one thing.

The cloud provider I use for my hybrid cloud setup is AWS. I choose eight out of ten for CloudDat because it handles more in a cloud way, and it uses AWS server and all that. There are other tools in the market where it handles different clouds. Advanced tools are also there in the market. I would rate this review as an eight out of ten overall.