Nada: Nodes-as-a-Service

A backend platform would maintain many horizontally-scalable pools of nodes. Access to these would be controlled via API keys & accounts.

I am a cloud DevOps engineer at Google, with a specialization in automation & enterprise scaling.

The platform will be API driven, with quota placed on accounts to throttle their requests per day. The service will be free up to a certain usage, after which it may charge for power users who perform a high-rate of requests.
By giving developers free access to nodes, they will be able to test their dApps & scripts without needing to administer a node themselves, nor deal with the time-consuming problem of syncing blockchain state.
User Experience
The design goal is to optimize fast usability, with no technical knowledge required by the end-user. The user is only tasked with creating an account at a website, upon which they will be given a personalized URL for their node's endpoint. Their local client will be able to interact with this node at this endpoint.
What this service provides for the user is the elimination of all tasks related to node maintenance. An application developer can simply focus on making apps. The toil of setting up & maintaining a node is taken care of by the Nada service.
The Nada web portal will give the user a dashboard about the usage of their node, letting them track request count & context.
Premium Users
To avoid bad actors, and keep costs under control, there will be throttling & quotas placed on users. The API will put a cap on:

• Requests per minute
• Requests during any sliding 24h period
  There will be generous limits in place for free users, and for entities that want high-frequency interactions with Nada nodes, a premium tier will be offered. The premium tier will be how this service monetizes itself, in order to stay up and running for free users.
  Backend implementation
  The backend of this system will make use of containerized Cardano nodes to accommodate auto-scaling and load-balancing. These techniques will ensure reliability and cost-effectiveness of the system.
  For connecting the Cardano CLI to the remote node, a small utility program will be written that forwards TCP traffic from the local socket onward. Or, for users who are not able to install the Cardano CLI, a small sandbox VM can be created for users to SSH into which will come enabled with the Cardano SDKs.
  Funding
  Funding will primarily be for developer time & infrastructure cost. As this service will attract users with a freemium model, it will need an early influx of capital to sustain the free users, until paid usage eventually balances the cost for operating hardware.
  20k - 400 hours engineering & DevOps work
  15k - Infrastructure for 1 year at a forecasted 1k free active users
  5k - Website, publicity, advertising

Code Licensing

Certain modules of the code will be open-source (MIT) license. These will be all smart contracts & "on-chain" logic. Critical systems that run on a centralized server will be closed-source for security reasons.