After several years of building and operating platforms on cloud, I started Metadash to fix the way things are done. I thought I’d share a few thoughts on where Metadash is going.
Cloud is too hard
Cloud was once easy. When it first started, there were just a few simple cloud services that helped users solve infrastructure problems with a couple of clicks. Since then, cloud has exploded into hundreds of services (with dozens of components each) across many providers.
Cloud became hard because it became complex. It was inevitable - cloud providers serve all kinds of users, from solo developers to global enterprises, across many industries and applications. No individual or team can really understand all of it anymore.
The end result is that every year over $55B is lost due to human error, manual work, and operational inefficiency. Cloud is just too hard, there are too many things to think about, too many opportunities for mistakes. Users assume since there’s nothing they can do, they must tolerate what exists.
Current solutions get it wrong
The way things are done now is users stack tools on top of tools to help them understand and optimize their cloud systems. One problem is that these tools sacrifice user control or user experience in order to do their job. Another problem is that most tools are passive and reactive, they can’t prevent problems or fix them.
For example, infrastructure-as-code trades experience for control. People don’t want to write or read code to manage infrastructure and even small deployments end up as pages of text. Cloud consoles have notoriously bad interfaces, but users prefer them if they can get the job done faster.
Ignoring the complex reality of the cloud is the wrong approach. Today’s solutions only add to the noise and ultimately don’t make the experience easy. They don’t remove unnecessary work, instead they just move it around.
There is a better way
The cloud experience has to be reworked from the ground up. There needs to be one simple interface for any cloud, rather than a different experience for every provider. It should be organized around primitives and workflows, not around services, since that’s how users solve problems.
Best practices should be integrated at every layer of the experience. Every component should have security, stability, and scalability built in from the start and be continuously optimized after deployment. Every change should have an audit trail and code output, so users have both visibility and control.
Most cloud operations should be automated, which is now possible with advances in AI - agents in particular. AI can carry out complex plans that would have been completely manual just a couple of years ago. The nature of large language models also lends itself nicely to solving problems in the cloud domain, where ambiguity is low and complexity is high.
What to expect ahead
The Metadash product consists of three parts: Next, Audit, and Agent.
Metadash Next is a simple and modern way to build cloud systems, starting with essential infrastructure. With Next, users can go from application code to a production-grade deployment in seconds by automating nearly all of the usual processes. After deploying, users can manage their system with full visibility and control over operations, while Metadash provides continuous monitoring and proactive remediation of cloud issues.
For existing deployments, Metadash Audit helps users find and fix critical security and compliance issues in their cloud systems. Mature organizations can use Audit to bring their infrastructure up to industry standards gradually and safely. Designed to handle complicated real world situations, Audit delivers results faster than the alternatives at a fraction of the cost.
Finally, Metadash Agent turns AI agents into infrastructure, leveraging Metadash Next to deploy production-grade agents easily. Agents have unique operational properties - unpredictable costs, heavy integrations, quality evaluations - that require complex configuration. Metadash Agent handles all the details so users can focus on making better products.
Building for the next wave
The next generation of companies will have cloud and AI at their core. The infrastructure they’ll use and the ways they’ll use it will be substantially different from previous companies. They need a platform that gives them all of the powerful capabilities of cloud and AI, without all of the accumulated operational complexity.
Metadash is on a mission to build that platform, applying lessons learned over years of production experience with infrastructure to make cloud easy. There is a lot to get done, but the path ahead is clear. The future of cloud is arriving very soon.
