Over the past few weeks, we’ve been refining our roadmap for 2026 and focusing on where Semaphore can deliver the most value.
Our direction is clear: extend CI/CD beyond execution — and help developers ship software faster.
Here’s what’s coming next.
Watch the quick overview by Marko Gracesa, our Head of Product:
Agentic Semaphore: The Next Phase of CI/CD
One of our major focus areas for 2026 is what we call Agentic Semaphore — extending CI/CD with AI-driven assistance while keeping developers fully in control.
CI/CD has always been about reliable execution. In 2026, we’re building on that foundation by reducing the friction around it.
We are currently working on two primary initiatives.
1. AI-Powered Onboarding
New users should not have to start from a blank YAML file.
We’re building an onboarding agent that:
- Converses with you about your project
- Asks clarifying questions
- Sets up a working Semaphore configuration
- Retrieves workflow digests and summaries
The goal is simple: drastically reduce the time between sign-up and your first successful pipeline.
As we prepare for broader campaign launches tied to our upcoming pricing changes, onboarding becomes critical. If we’re attracting new developers, the experience must be fast, intuitive, and frictionless.
Instead of reading documentation and configuring everything manually, developers define intent — and Semaphore translates that intent into a working pipeline.
2. Faster Root Cause Detection for Failures
For existing users, one of the most important questions in CI/CD is straightforward:
Why did my pipeline fail?
We’re iterating on AI-assisted failure analysis that can:
- Digest workflow results
- Identify likely root causes
- Summarize what happened
- Help you move from red to green faster
The focus is speed and clarity. When a workflow fails, the time between failure and resolution should be as short as possible.
Unified Storage
The Unified Storage project has officially been approved after being evaluated against our four key criteria: valuable, usable, feasible, and viable. The answer to all four is yes.
From a customer perspective, the most significant improvement will be cache performance — expected to be up to six times faster than today.
This will have a particularly strong impact on workflows with multiple test jobs. Currently, when several jobs need to pull the same cached data, each one downloads it independently. With unified storage, those uploads and downloads will happen much faster, significantly reducing workflow duration.
We are also improving how we handle the “noisy neighbor” problem, ensuring performance is managed more gracefully so one workload does not negatively impact others.
Overall, this means shorter workflows and faster feedback loops.
Internally, unified storage brings major operational benefits as well. Scaling storage becomes significantly simpler. Instead of manual intervention, we can add storage capacity seamlessly. The system is also more enterprise-grade in terms of user management and platform integration.
The project benefits both customers and our internal operations: users get faster pipelines, and we reduce operational overhead.
If everything stays on schedule, execution begins around February 22, with a target availability by the end of Q1, pending final engineering confirmation.
GitLab Integration Nearing GA
As previously shared, our GitLab integration is nearing General Availability.
This includes:
- Sign up with GitLab
- Connect GitLab repositories directly
- Streamlined CI/CD setup for GitLab workflows
We are intentionally aligning this launch with our upcoming pricing announcement to deliver maximum value for the GitLab community.
This integration reduces friction and makes Semaphore more accessible to teams where their code already lives.
Our Roadmap Forward
CI/CD remains the foundation of Semaphore.
But our roadmap is focused on extending that foundation in meaningful ways:
- Making onboarding dramatically easier
- Reducing time spent diagnosing failures
- Embedding assistive AI into developer workflows
- Improving infrastructure performance at scale
- Simplifying pricing to remove feature-based barriers
We are not pursuing AI for its own sake. We are using it to remove developer toil.
We are not adding complexity. We are reducing friction.
Every initiative we move forward must be valuable, usable, feasible, and viable. If it doesn’t meet all four criteria, it doesn’t ship.
In the Coming Weeks
You can expect:
- Easier onboarding
- Smarter workflow insight
- Unified Storage rollout progress
- GitLab General Availability
- A simpler pricing model
All in service of one goal:
To help developers ship software faster.
If you haven’t tried Semaphore yet, you can sign up and start shipping today.
More updates soon. 🚀
Want to discuss this article? Join our Discord.
