Platform Engineering - Emergence AI: Always fly (an IDP) with a copilot

The post-DevOps moment

The promise of platform engineering was to make cloud-native development less complex, increase developer productivity and make shipping software more predictable, explaining its focus on “golden paths” and internal developer platforms.

This led to a surge of large software engineering organisations adopting it. Gartner predicts that 80% of them will have a platform engineering team by next year.

The path, though, is not entirely smooth. At many companies, adoption of IDP frameworks like Backstage struggle to rise above 10%. A core reason for developer hesitation is that, as with any platform, an IDP has its own learning curve and complexity and at runtime, its own challenges with visibility and debuggability.

Getting past this adoption hump is a key place where AI can help now. Millions of developers have ingrained AI into their daily workflow with tools such as GitHub Copilot, Cursor, Claude Code, OpenAI Codex and others. AI copilots and agents have boosted developer productivity and markedly reduced time from idea to implementation. Now we can apply this same approach to platform engineering. We can bring the power of copilots and agents to overcome IDP adoption challenges.

Beyond the adoption hump

Today’s IDPs are feats of engineering and abstraction. They centralise service catalogues, automate infrastructure and standardise CI/CD pipelines. But developers on the ground still encounter walls of configuration files, endless documentation and cryptic error messages pouring out of build logs that make these platforms feel unnecessarily complex. IDPs offer a menu of options, but the developer still needs to understand them to choose the right one for the particular problem they’re solving. 

An IDP still cannot answer the most fundamental developer questions: “Why did this break?” “What’s the right way to do this here?” or “What do I even need to ask for?”

The IDP vision

The next generation of IDPs will answer those questions by evolving from a passive menu of options into an active, intelligent partner. If a personal copilot in your editor helps you with your code, the Platform Copilot will help you with everything else, from code structure and design, to deployment, observability and debugging. It will allow a developer to state their intent in natural language and then the platform’s job will be to understand that intent and translate it into action.

This leap is made possible by connecting a “brain” to a “nervous system” so that the IDP, with its rich set of APIs, service catalogues and observability data, is the platform’s nervous system. It has the means to act and to sense the state of the entire system. The brain is a Large Language Model (LLM), providing the reasoning engine to understand requests, analyse data and formulate a plan.

The fusion of the two gives rise to a new, intuitive way of working called “vibe coding” for your internal workflows. A developer no longer needs to write a precise pipeline definition; they can simply describe it. 

The platform copilot in action

Let’s make it tangible with a few examples.

Haldar: The measure of an advanced platform will no longer be the slickness of its UI, but the quality of its conversation.

For the internal developer (platform engineer): A new engineer joins the team and asks the platform, “Hey, how do I build a new Go microservice that talks to a Redis cache?” The copilot scaffolds the service using the company’s approved templates, creates the repository, generates the initial CI/CD pipeline and links to the relevant security checklists. Later, if a deployment fails, the developer can ask, “Why did my deploy to staging die?” The copilot, using its AIOps foundation, analyses logs and metrics, identifies a missing network policy as the root cause and suggests the correctly formatted code to fix it.

For the site reliability engineer (SRE): An alert fires for high latency on the payment service. The copilot automatically enriches it with crucial context: “Latency is spiking. This correlates with a new deployment (commit #a4d8c3) and a surge in traffic from the promotions API. Would you like me to initiate a rollback to the last known good version?” It also works proactively, flagging that “the staging database for the now-merged Project-X has been idle for 14 days, costing an estimated $200/month. Shall I decommission it?”

For the product manager: A non-technical stakeholder can finally get answers without deciphering complex dashboards. They can simply ask, “What’s the health of the mobile login feature?” or “Which services are currently running in the UAT environment for the Q3 release?”

The future is conversational

This is not science fiction.

From our vantage point at Emergence AI, we see the challenge of creating a Platform Copilot is similar to building any advanced AI agent: one must connect a powerful language model to a well-defined set of tools and data sources. The IDP provides the perfect “body” for an AI “brain” to inhabit, ready to translate human intent into machine action.

The evolution is here.

The measure of a truly advanced platform will no longer be the slickness of its UI, but the quality of its conversation.

It’s a fundamental shift in our relationship with the tools that power our digital world. The era of describing your needs and having an intelligent platform partner and bring them to life has arrived.