Skip to content
Nubent

/Platform engineering

Platform engineering: your internal developer platform on AWS

We design and build internal developer platforms on AWS so your teams can create services, environments and infrastructure through self-service, on secure standard paths and without waiting on tickets. Golden paths, a service catalog with Backstage, reusable Terraform modules and automated guardrails.

/Diagnosis

Signs it is time

Infrastructure is the bottleneck

Every new service, database or environment needs a ticket to the platform team, and the wait slows down the rest of the organization.

Every team does it differently

There are as many ways to deploy, log and configure as there are teams, and keeping security and consistency becomes impossible.

Slow developer onboarding

A new hire takes weeks to figure out how to create a service, where the dashboards are and who to ask for permissions.

Copy-pasted Terraform

Duplicate modules with small differences live in every repository, and a security fix has to be applied by hand dozens of times.

/What we deliver

Concrete deliverables, in your repository

  • Platform strategy

    Team interviews, a friction map, priority capabilities and a roadmap for the platform, treated as an internal product.

  • Golden paths

    Service templates with repository, pipeline, infrastructure, observability and security ready to go, for your most common stacks.

  • Backstage portal

    A service catalog with owners, technical documentation and project scaffolding, all from an internal developer portal.

  • Reusable Terraform modules

    A versioned library of modules for databases, queues, services and networking, with security policies and tags built in.

  • Self-service with guardrails

    Flows to create environments and resources without tickets, with policy validation, automated review and cost limits.

/How we work

Our process

  1. Developer research

    We measure where teams lose time, which tasks repeat and which standards are missing, using interviews and delivery data.

  2. Minimum viable platform

    We build the first complete golden path for a pilot team, from scaffolding all the way to production.

  3. Adoption and expansion

    We move more services onto the standard path, add self-service capabilities and adjust based on team feedback.

  4. Running it as a product

    We track adoption and delivery metrics, maintain the modules and hand operations over to your platform team.

What platform engineering is and what problem it solves

Platform engineering is the discipline of building and running an internal platform that development teams use as a product. Its goal is to reduce cognitive load: instead of every developer learning the details of IAM, networking, Terraform and observability, the platform offers standard paths that already make those decisions safely.

It emerged as an answer to a specific problem with DevOps at scale. Asking every team to run its own infrastructure works with a few teams, but with more services it breeds inconsistency, duplication and risk. Centralizing everything in an infrastructure team that works through tickets creates bottlenecks. An internal platform balances autonomy and standards.

Golden paths: make the easy way the right way

A golden path is a recommended, supported way to build a type of service, for example a containerized API or a worker that consumes a queue. It includes the repository template, the pipeline, the infrastructure modules, log and metric configuration, default alarms and documentation, all ready to use.

The standard path has to be more convenient than leaving it. If creating a service with the golden path takes minutes and doing it by hand takes days, teams adopt it without anyone forcing them. We start with the two or three most common service types in your organization and expand based on real demand.

Backstage and Terraform self-service on AWS

Backstage, the CNCF open source project, works as a developer portal: a catalog of every service with its owner, documentation, dependencies and dashboards, plus templates for creating new projects. We integrate it with your repositories, pipelines and AWS accounts so it becomes the single front door.

Behind the portal, self-service runs on versioned Terraform modules published to a private registry. A developer requests a database or an environment, a pull request is generated with the right module, policies are validated with tools like Checkov or Open Policy Agent, and the change is applied automatically. For simpler catalogs, AWS Service Catalog is also a valid option.

How to measure the success of an internal platform

An internal platform is a product, and products get measured. We use delivery metrics such as DORA, which cover deployment frequency, lead time for changes, change failure rate and time to restore, alongside adoption metrics like the share of services on golden paths and the time it takes to create a new service.

We also measure developer experience with short, regular surveys. If the platform does not reduce friction, or teams route around it, it needs to change. That is why we work in short cycles with a pilot team rather than a big upfront design that takes months to deliver value.

/FAQ

Platform Engineering: frequently asked questions

Another question? Write to hello@nubent.com and an engineer will answer.

What is an internal developer platform (IDP)?

An internal developer platform is the set of tools, templates, automations and portals a platform team offers the rest of the organization to build, deploy and run software through self-service. It packages security, infrastructure and observability best practices so each team does not have to solve them on its own.

What is the difference between DevOps and platform engineering?

DevOps is a culture and a set of practices that help development and operations collaborate and ship software frequently and reliably. Platform engineering is a way to put those practices into effect at scale, by building an internal platform as a product that makes it easy for teams to follow the right path without relying on tickets.

Do you need Backstage to have an internal platform?

No. Backstage is a popular choice for the portal and catalog, but an internal platform can start with repository templates, Terraform modules and standard pipelines. We recommend a portal once there are enough services and teams for a catalog and centralized documentation to add real value.

At what size does platform engineering become worth the investment?

There is no fixed number, but the signals usually appear when several product teams deploy services independently and the infrastructure team starts to become a bottleneck. Before that, good standards and shared modules are enough. A minimum viable platform lets you start without a large upfront investment.

How is security enforced in a self-service platform?

With automated guardrails instead of manual reviews: Terraform modules that already include encryption, tags and least-privilege permissions, policy as code that blocks risky configurations at the pull request, organization-level SCPs and continuous monitoring with Security Hub. Developers gain autonomy within safe limits.

Next step

Let's talk about your Platform Engineering project

30 minutes with an engineer. Concrete next steps, whether or not you work with us.