Hi guys, here at Quave we are still building strongly with Meteor especially for larger companies, recently we rebuilt 2 entire companies’ web stacks in Meteor 3 and it’s doing great.
One is the biggest data provider of Ecommerce Data in Brazil (1.5M transactions a day) and the other is one of the biggest return platforms for Ecommerces, powering customers like Samsung and other huge ecommerce brands.
We’ve been asked sometimes about running our cloud on AWS and that was always an option but we didn’t have a good name and way to explain, now we are fixing that. We have Mario onboard, former fintech/education Founder as our CRO and we are going to the market explaining what we do, finally.
Moving forward we will call our cloud platform Quave ONE.
I’ll keep this practical and straightforward, let’s talk about how we solve the issue of running web apps, APIs, databases in any cloud anywhere in the World and of course, also on AWS.
What Quave ONE is, in simple terms
Quave ONE is a cloud-agnostic platform that sits on top of your infrastructure and handles the operational side of running apps and databases.
With it, you can:
- Run Meteor apps inside your own AWS account, keeping your VPC, regions, provider credits, and setup
- Use GCP, Azure, OCI, on-prem, or even multiple providers if you prefer
- Avoid long-term lock-in to a single cloud
- Deploy, scale, migrate, and roll back with one click and without downtime
When you run in your own account, on your own provider we call it Quave ONE - Bring Your Provider.
If you don’t have a strong preference for a provider and just want things to be easy, there’s also Quave ONE Direct, where apps run on Quave-chosen infrastructure. No cloud provider account setup.
In both cases you don’t do anything, everything is handled by us. We have cases of Meteor apps saving 90% and running 6x faster compared to their previous cloud solution, check our case studies to learn more.
Meteor apps for obvious reasons are very popular in our cloud, I would say we have probably more than 500 Meteor apps and more than 100 MongoDB clusters for sure (I haven’t checked our database to post exact numbers, these are safe guesses).
MCP: working with infra and logs using natural language
We recently launched MCP (Model Context Protocol).
MCP lets you interact with your infrastructure using natural language. It has access to infra state, metrics, logs, deployments, and costs. You control the permissions: read-only, suggest actions, or execute after approval.
This has been especially useful for debugging and log analysis. A few concrete examples:
-
“My Meteor app is slow right now. What’s going on?” - MCP analyzes logs and metrics and points to concrete causes like memory pressure, or CPU/memory issues, etc.
-
“Show me all errors and warnings from the last 30 minutes and group them by root cause.” - Instead of manually digging through logs across services.
-
“We’re seeing timeouts. Is this coming from the app, the database, or networking?” - MCP correlates signals across layers and narrows it down.
-
“Scale the API layer for the next traffic spike.” - MCP proposes a plan and only applies it if you approve.
And this is just the beginning, we are adding new MCP Tools weekly and also providing more details to existing tools, just this weekend we released a tool on our MCP so it knows how to read our docs, for example.
Docs for our MCP are here if you want to dig deeper: MCP (Model Context Protocol) | Quave Cloud Docs
Other platform basics
- One-click deploys
- Zero-downtime deploys
- Easy migrations between regions or providers
- Observability out of the box (Prometheus + Grafana)
- Backups, rollbacks, and simple scaling or advanced autoscaling, for example, in your account on AWS we can run in multiple regions with autoscaling of VMs and also autoscaling of Apps.
It’s amazing, we have a Meteor company running 4 AWS regions this way with us. Zero setup on their side. 100% managed and migrated by us. They migrated from running directly on AWS to us running on their AWS account to have all the benefits of Quave ONE and they are really happy and saving money. How do they save money? When you run on AWS directly you pay for every layer of convenience you add to your stack, think about Fargate, with us, you run just simple EC2 instances and the LB of AWS if you want, everything is with us, our technology and then you don’t pay a premium on top of everything.
About support (the human kind)
One thing that matters a lot for teams running Meteor in production is real human support.
With Quave ONE, you get access to engineers in minutes, not days. Real people. Curious nerds who understand Meteor internals - many of them former core members, MongoDB, and cloud infrastructure (AWS included).
The goal isn’t just to fix tickets. It’s to help you reason about production issues, make better infrastructure decisions, and scale things in a way that actually makes sense for your app and traffic patterns.
You can run everything self-service if you want, especially with MCP. But when production gets weird, having someone experienced to talk to makes a big difference.
BTW, 95%+ of our support tickets are not about infrastructure, we help clients with code issues, db issues, anything you want, even helping you to decide the best stack pieces for your new project and how to save on costs with self-hosted solutions, we run everything, from simple apps to highly customized set ups.
For those who remember the times I was running Meteor Cloud, you probably remember me as I was responding to most of the tickets all the way with deep analysis helping our customers to succeed.
That’s it
Not here to push anyone to switch, just sharing Quave ONE as an option for teams that want to keep running Meteor on AWS (or stay flexible if things change later), we’ve been doing these things for more than 2 years but we were not speaking about it publicly because we were busy doing things for customers, but now with Mario, this changes.
If you have questions, want to talk through a scenario, or just want to sanity-check your options, feel free to reach out: filipe@quave.com.br or schedule a call directly (busy holiday times coming) https://cal.com/filipenevola/cloud
Also, about pricing, the Quave ONE Bring Your Own Provider plan doesn’t have dynamic prices based on hourly usage, you pay us a fixed license fee monthly, it starts at $399 USD. Great for companies spending $2k or more in total cloud + devops costs. Remember that your devops team and costs will be reduced or even zero when running with us (sorry devops guys - ping us if you want to work with us).
Or in the Direct plan you pay 7.5 USD per 1 zcloud (0.5 GB of RAM and 0.5 vCPU) - they are very powerful CPUs, so you are probably going to cut by half or more what you have today.
If you need to decide your next provider we also help, and even negotiate credits and special plans for you with cloud providers and we are also Official Oracle Partners, if you don’t know anything about OCI please ping us, it has the best cost-benefit from all big clouds by far.
Happy to chat.