LLM Gateway
TrueFoundryReviews from AWS customer
0 AWS reviews
-
5 star0
-
4 star0
-
3 star0
-
2 star0
-
1 star0
External reviews
51 reviews
from
External reviews are not included in the AWS star rating for the product.
Kubernetes made easy
What do you like best about the product?
I was looking for ways to deploy our docker images to serve AI models. I first tried deploying manually on a VM, and through Azure Container Instances. But I always found it hard to manage the continuous deployment of our different services for our different environments (dev/staging/prod). And we always provisioned more resources than were necessary, because of the difficulty to auto-scale, and re-start the instances deployed on spot instances (for dev & staging).
TrueFoundry solved all our deployment-related issues. It was very easy to get our first application deployed, thanks to their good default settings and clear GUI. And, as you get more familiar with it, you start clicking the "show advanced fields" or configure by directly editing YAML files. So, it's easy for onboarding, while allowing power users to use advanced features.
At first, our CEO was not enthusiastic about paying for yet another service. But it was easy to convince him, when we showed him the computing costs savings we achieved as a result of adopting TrueFoundry, were greater than the cost of the service (and that's without counting the time and headaches it saves us). Sure, we might have achieved the same computing cost savings by managing the kubernetes clusters ourselves, but the engineering costs would have been much much higher.
But what I like the best about TrueFoundry is their active customer support. Not only have they answered all our questions precisely and rapidly, they updated the product based on our feedback (e.g. making it easier to deploy frequently used applications such as LabelStudio), and they double checked our Kubernetes configuration and suggested improvements.
TrueFoundry solved all our deployment-related issues. It was very easy to get our first application deployed, thanks to their good default settings and clear GUI. And, as you get more familiar with it, you start clicking the "show advanced fields" or configure by directly editing YAML files. So, it's easy for onboarding, while allowing power users to use advanced features.
At first, our CEO was not enthusiastic about paying for yet another service. But it was easy to convince him, when we showed him the computing costs savings we achieved as a result of adopting TrueFoundry, were greater than the cost of the service (and that's without counting the time and headaches it saves us). Sure, we might have achieved the same computing cost savings by managing the kubernetes clusters ourselves, but the engineering costs would have been much much higher.
But what I like the best about TrueFoundry is their active customer support. Not only have they answered all our questions precisely and rapidly, they updated the product based on our feedback (e.g. making it easier to deploy frequently used applications such as LabelStudio), and they double checked our Kubernetes configuration and suggested improvements.
What do you dislike about the product?
They are still new, so some parts of the user interface could be improved. But it is already quite good, and I have seen improvements over the last few months.
What problems is the product solving and how is that benefiting you?
It allows our ML engineer (which are rarely knowledgeable about those kinds of things) to deploy and monitor their applications themselves. This avoids the anti-pattern of ML teams throwing their models to a separate Ops team. But it still allows fine-grain control, when an Ops specialist needs to jump in.
It made it easy for us to implement Continuous Delivery (we auto-deploy to our dev environment every time a push on origin/main changes a transitive dependency of a service).
It allowed us to reduce our computing costs in several way:
- It made it easier to use spot instances for our dev and staging environment, reducing the costs of those environments by 80%.
- It made it easy to turn off our most expensive services during nights and weekend (we only need them during office hours).
- It reduced resource waste, by making it easier to run multiple small services on the same machine in an isolated way.
It still feels we are under-using True Foundry, since we are mostly deploying "services" and haven't deployed jobs or notebooks yet.
It made it easy for us to implement Continuous Delivery (we auto-deploy to our dev environment every time a push on origin/main changes a transitive dependency of a service).
It allowed us to reduce our computing costs in several way:
- It made it easier to use spot instances for our dev and staging environment, reducing the costs of those environments by 80%.
- It made it easy to turn off our most expensive services during nights and weekend (we only need them during office hours).
- It reduced resource waste, by making it easier to run multiple small services on the same machine in an isolated way.
It still feels we are under-using True Foundry, since we are mostly deploying "services" and haven't deployed jobs or notebooks yet.
A promising exotic wrapper around cloud providers
What do you like best about the product?
Its inherently built around, reliability, cost saving, but also gives freedom in easily use or custome services as required, be it for deployment or for development. UI is very intuitive, customer support is top notch. Supports large number of integrations from code to direct docker images.
What do you dislike about the product?
TrueFoundry though fast pacing, still lacks some basic features that might be holding it back to deploy in scale.
What problems is the product solving and how is that benefiting you?
Deploying multiple services, the instance selection or pooling, exposing the ports and no downtime was immensely helpful
Great platform to deploy ML system easy at scale
What do you like best about the product?
- Expertise support
- User intuitive platform
- Best for Service/Model deployment
- Featuristic dashboard for model management
- Ease of use for model traning and fine tuning
- User intuitive platform
- Best for Service/Model deployment
- Featuristic dashboard for model management
- Ease of use for model traning and fine tuning
What do you dislike about the product?
Not seen any major cons with TrueFoundry. Since an early product minor issues with Hugging face open source models deployment.
What problems is the product solving and how is that benefiting you?
- Easy model deployment
- Great support for model traning and fine tuning
- Great support for model traning and fine tuning
Really useful
What do you like best about the product?
makes launching k8s a million times easier with really great support from their team
What do you dislike about the product?
just some ui things could be better but cant think of anything else
What problems is the product solving and how is that benefiting you?
deploying lots of inference endpoints
Deploy models easily on any cloud machine with inbuilt autoscaling
What do you like best about the product?
- Autoscaling
- Cloud Agnostic
- Customer Support
- Easy to use and very easy to integrate in existing code unlike AWS Sagemaker or Metaflow which requires chaning code
- Once setup is done, easy to update model with a single click
- Cloud Agnostic
- Customer Support
- Easy to use and very easy to integrate in existing code unlike AWS Sagemaker or Metaflow which requires chaning code
- Once setup is done, easy to update model with a single click
What do you dislike about the product?
- Since it is an early product, few featuers might not be available but customer support is very good and they implement features at very fast pace
What problems is the product solving and how is that benefiting you?
Easily deploy ML models which is very complex on AWS Sagemaker or Metaflow.
Existing solutions look very intimidating and require a lot of code changes whereas True Foundry just acts as a layer on the infrastructure which makes it easy to use and integrate.
We deploy ML models for users and the infra needs autoscaling on peak time. True Foundry provides that out of the box which solves the major problem of setting up own Kubernetes based machine.
Existing solutions look very intimidating and require a lot of code changes whereas True Foundry just acts as a layer on the infrastructure which makes it easy to use and integrate.
We deploy ML models for users and the infra needs autoscaling on peak time. True Foundry provides that out of the box which solves the major problem of setting up own Kubernetes based machine.
A one stop DevOps solution
What do you like best about the product?
I find TrueFoundry to be a great solution to solving common DevOps related problems and setups. They have reduced the time to deploy applications significantly and made it much easier and user friendly for anyone to use. They also offer great support when we end up facing possible issues.
What do you dislike about the product?
I haven't found any major downsides or cons with the product. But the documentation can possibly improve with pointers to commonly faced problems.
What problems is the product solving and how is that benefiting you?
TrueFoundry is helping us solve a lot of DevOps problems that would require significant effort from the team.
A great product to quickly deploy ML system at scale.
What do you like best about the product?
The support team is quick to solve any problems. Migrating your existing solution is easy.
What do you dislike about the product?
Nothing major. Logging can be improved.
What problems is the product solving and how is that benefiting you?
We wanted to deploy our system to work at scale. For this, we wanted to create an asynchronous autoscaling service that can scale GPU instances based on the load. We initially tried Sagemaker, but found TrueFoundry to be much easier and reliable.
Best learning platform
What do you like best about the product?
The uses cases of LLM training modules..
What do you dislike about the product?
Nothing I could see and everything good.
What problems is the product solving and how is that benefiting you?
Text Extraction and Entity Extraction .
Great platform to deploy micro-services & servers in a minute
What do you like best about the product?
I like the easy-to-use UI from which you can directly deploy, and monitor your microservices. It also abstracts out a lot of complications that might be required to set up things like Canary rollouts, A/B testing, etc.
TrueFoundry abstracts all of the complications of managing Kubernetes for us while optimizing our cloud costs
TrueFoundry abstracts all of the complications of managing Kubernetes for us while optimizing our cloud costs
What do you dislike about the product?
As I mainly deploy the services, using the UI, it's quite a hassle to set the configuration settings, again and again
Apart from this, the UI seems to be working well for me
Apart from this, the UI seems to be working well for me
What problems is the product solving and how is that benefiting you?
Thanks to truefoundry, the overall time to deploy our services & servers has reduced significantly!
Sophisticated mlops tool set for machine learning model tracking and deployment
What do you like best about the product?
With their sophisticated MLops tool set, we quickly resolved our model deployment and tracking worries in no time. It's so flexible that users can easily manage the deployed solutions through their centralized dashboard and also easy to work between users and teams. The user can get diverged view from their interactive plots and one-to-one comparison between the models as part of model tracking.
What do you dislike about the product?
Nothing to be specific to dislike in this tool set. It serves the purpose and also enhances the way the developer will solutionize the model tracking and eployments
What problems is the product solving and how is that benefiting you?
Before the TrueFoundry MLops toolset, we are spending hours deploying our machine learning model and getting them to serve the live traffic. But this toolset have saved almost ~60% of our time on deployment-related activities. Also, with efficient model tracking, we could quickly call on our star solution.
showing 41 - 50