At Railway, we take customer feedback very seriously. We structure most of our communications around this principle. For example, we pipe all emails to our chat channels, consistently acknowledge and prioritize user input, and try our best to set expectations when highly anticipated features get discussed.

We feel the product and our company benefits when our users believe in our vision: allowing you to work on your core product without worrying about infrastructure. The constant tension for any company is when the roadmap of highly anticipated features rubs against the reality of the essential needs of your users.

The Railway platform is constantly evolving to meet users' needs, large and small. Even though we think we do a great job of letting everyone know what we are thinking about in the Changelog, we can always do better. Hence, the idea for our first town hall-like event was born. Our first call was on our community server using Discord stages. We fielded questions using Sli.do up to a week in advance. We feel the actual event went very well. Although we cannot provide a recording for this call, we thought it would be prudent to share answers to common community questions related to our roadmap.

Before we jump into the questions, let us share a bit of our near-term roadmap.

For many users who have been on the platform for a while, after we GAed the new project canvas, we have been hard at what we call "grease work." We have immediately eliminated an entire class of outages during the last month while polishing up typical user flows on the dashboard.

Our infra is now much more declarative, where we can scale to meet user demand much more gracefully than before. However, with all the improvements we made, it's becoming increasingly clear that we need to place much more work on our "bones" of the platform to support the more significant features that companies need. A good example is our much-discussed plans for private networking. Networking is something we think we can get right with an excellent user experience but requires a lot of plumbing within our back-end to enable customers on the platform.

We have dubbed this effort "Infra v2". As we work to solidify and strengthen our infrastructure for today, we also look onward to tomorrow to build out and scale up to handle the massively growing needs of the market. The next version of our infra is essential work that will help us enable the next generation of the platform.

The work to come was a theme referenced quite a bit during our call.

We fielded around 40+ questions from an engaged audience during the call, and we thoroughly enjoyed the discussion. This is our best attempt to summarize and answer the questions asked for those who could not attend.

What is the status of Railway's API? When can we expect a beta for that, and how will it work?

Reasonably solid but on the back burner as Nixpacks is taking priority. API potentially will be within the next month, but probably just a team feature.

We expect the API to allow users to query information about your projects and will enable you to set some settings programmatically. The API will be delivered via graphQL.

When will Nixpacks release for users?

It will be available for Priority Boarding users as of this post. We will make it available to everyone on Railway when it is relatively stable (compared to the Heroku and Paketo buildpacks). We assume there will be many issues that come up when testing on larger apps. So we will focus on stability over the next few weeks.

Is Remote Shell on your roadmap? Especially for running `rails console`

We know this is a sticking point for many Ruby and elixir developers. We will not support SSH in your application because it breaks the immutability principle when deploying your applications.

What is the status of GitLab integration?

Planned but no timeline. Much of the product is closely tied to GitHub's deployment flow and model. A lot of work will need to be done to disintermediate this.

Could serverless support be added to Railway?

No plans at this time for serverless or "scale to zero." We suggest using other vendors such as Vercel. We have an integration if you wish to augment your applications.

Is Railway's Database options production-ready?

At the risk of sounding like a non-answer... Production standards are going to be different depending on what your users expect. Let's start with what our DBs don't have: no SLAs, no HA, and it can only go so far vertically. We don't think they are suitable for anything mission-critical, like if you wanted to start a bank. With that said, we do have customers using our DBs for their production environment with no issue. We don't believe in vendor lock-in here at Railway, so if your needs outpace us, consider other vendors like PlanetScale (for MySQL) or Cockroach (For Postgres).

What is the status of custom plugins/databases? When will that be available?

We want to dogfood the ability to author interfaces for those integrations as much as possible. But unfortunately, custom anything on the platform is a pandora's box and will take a while. Figma famously took six years to release custom plugins, although we don't think we will take that long.

Are there any updates regarding Postgres modules?

Not planned at this moment. Will prioritize foundational infra work.

When will persistent storage be available?

We're well aware that people want persistent storage. Unfortunately, stateful orchestration is a very hairy problem that opens up an entire system of issues on its own. We'd like to. Short term, as part of Infra v2 supporting something like SQLite.

What are Railway's plans for dedicated IPs and private networking within projects?

As mentioned, we'll work on v2 of the infrastructure. We aim to include the ability to attach NATs/IPs to this. We plan this will be a paid/teams-only feature.

In the long term, we envision users grouping services within private networks and allowing people to set up firewall rules to control permitted traffic within their projects.

Are managed Postgres backups on your roadmap?

Planned as part of infra v2. It will require a plugin system rewrite but this is one of our most requested features.

When will SOC2/HIPPA compliance occur?

We won't get certifications anytime soon. The long answer is we have several fundamental platform changes incoming (multi-region, additional scaling solutions) that would affect our audits. If we were to undergo certification, we would have to amend the certifications quickly.

Any plans for Railway to have an on-prem version?

On-prem is in our long-term vision. However, we are ways off from being ready to cut releases for people to run their versions of Railway on their infra.

In closing, our community's desire to see specific use-cases fulfilled gives us the drive to write better requirements and enable our users to build applications that their friends and customers enjoy.

We have many things in the works to help us keep these communication lines open. Chiefly being, we will work on making a public page for our roadmap and scheduling more public calls to have burning questions answered. For teams, any immediate questions about incoming features can always be addressed in your Direct Support channel on Discord (or emailing us).

Thank you all for believing in our product, team, and vision. We hope to make your projects succeed today and every day after that.