Introduction

Welcome to the Omnistrate developer documentation.

Omnistrate is a developer platform, based on open standards, that turns any software into an enterprise-grade, fully managed, multi-cloud SaaS in minutes. Let's say you are a PostgreSQL developer and want to build a multi-cloud PostgreSQL SaaS, our platform can help you make that transformation while delivering a reliable and high-performance SaaS to your end-users.

In other words, we are a Control Plane as-a-Service that helps you build next the Snowflake, MongoDB Atlas, Databricks Cloud, Redis Cloud by live generating the control plane for your software in 1/10th of the time and cost.

Who is it for?

Teams who are building their SaaS or Internal platform teams who are trying to standardize and automate the undifferentiated work so that you can focus on the differentiated things. To be more specific, here are some of the common use-cases:

• If you are building a SaaS for your data software stack including one or more software components like databases, caching, streaming, search, analytics, messaging, data pipelines or any data processing application
• If you are trying to automate your platform engineering to operate efficiently, stay nimble to try out new technologies (particularly open-source), centrally define controls / practices.
• If you are migrating across clouds and looking for ways to accelerate your cloud migration journey

Why is building SaaS so painful and time-consuming?

To transform a software into SaaS requires several steps that can take years to get it right. The reason it takes so much time and effort is that cloud abstractions are too low level to build SaaS and are not compatible across clouds.

Here are some sample problems that you have to solve:

• As a SaaS offering, your customers will expect an automated, dedicated deployment stack. Terraform is a good starting point for one-off deployments but scales pretty poorly for such automation for every customer. Beyond the initial setup, customers will iterate over the configuration of their cluster, be it software versions or hardware configurations, and there is no way to automate this in a matter of minutes currently.
• How will your customers interact with your SaaS service?
• What about scaling your infrastructure when the load increases and scaling it down to zero when the load decreases?
• What about handling failures and providing an end-to-end visbility across the stack when the cloud infrastructure fails?
• What about billing your customers automatically when they use your services?
• What about security and compliance for enterprise customers including but not limited to authentication, RBAC, auditing, secure networking, infrastructure security from network segmentation, end-to-end encryption, vulnerability scanning?
• What about cloud-native capabilities that are tablestakes these days for any PaaS/SaaS service like observability pipelines, common actions like cluster start/stop, reboot, failover etc?
• How will you manage your fleet, gain insights into the fleet and handle common actions like patching CVEs or rotating certificates?

This can get pretty hairy pretty fast and this is just the bare-minimum necessary to be competitive in today's market. The problem is that it takes more than a year for engineering teams to build this basic SaaS offering only to realize the problem multiplies with multiple environments for continuous delivery, multiple clouds, multiple regions, different delivery models, different tenancy models, different networking models, adding new services and will highly likely end up building most of the stack again. From our experience, a significant portion of the engineering team is caught up in building the common undifferentiated platform as opposed to focussing on the core differentiation for your software.

For more details, please see this

How are we addressing this challenge?

Existing SaaS providers have built their own control plane layer but the problem is that they are custom built for their technologies and restricted to their cloud. Omnistrate aims to democratize this layer and make it easier to deploy, scale, and operate applications across multiple cloud providers and across multiple services. Here are some of the specific problems that we solve:

• Automated Infrastructure Provisioning

  • API/UI/CLI ready for your customers to perform different management operations on your SaaS like deploy, update, scale, reboot, failover, etc.
  • Get fully managed provisioning for all of your customers without managing Terraform state for every customer
  • In addition to supporting deployments in your account (hosted), you can enable Databricks style BYOA deployments in your customer's account reliably and securely.

• Continuous Deployment

  • Seamlessly deploy, configure, patch your software through automated deployments
  • Get full visibility into your deployments and resources on your service dashboard

• Automated CloudOps

  • Handle failure scenarios to help achieve 99.99 SLA from simple machine failures to complex AZ failures, network partitions, process deadlatches, gray storage nodes etc.
  • Manage your entire fleet from rotating certificates, upgrading infrastructure to leverage new instance types or storage types or more efficient CPU variants like Graviton, evolving your customer experience, getting insights into your fleet, metadata governance and much more
  • Get L1 support spanning from automated operations to in-depth metrics, logging, alerting capabilities using your favorite tools
  • Achieve compliance requirements for your enterprise customers by virtue of the underlying platform

• SaaS Tablestakes

  • Automatically enable cloud-native capabilities like autoscaling, RBAC, API keys, observability, auditing, secure networking out of the box
  • Make your software serverless scaling all the way down to zero
  • Complete your SaaS offering with integrated billing for your customers
  • Seamlessly deliver SaaS across multiple environments, clouds, regions, and deployment models with our platform

We have been working in the cloud industry since 2007 and have built dozens of SaaS services at AWS, Confluent, Microsoft, Oracle from the ground up and everytime we needed a management layer that provisions, scales and manages the cloud infrastructure for the underlying software to run. The problem is that everytime we ended up taking a meaningful portion of the engineering resources to build this layer as opposed to focussing on the core innovation.

Our vision is to make it a few clicks for anyone to build their SaaS and remove the pain of building the common management (aka control plane) layer by abstracting away the undifferentiated pieces into a central platform. With Omnistrate, we have developed a platform from the ground up using open standards to allow you to build, iterate, operate your SaaS offering quickly and efficiently.

For more discussions, please join our community here