Introducing Port AI assistant

Overview

We’re excited to introduce Port AI Assistant - your intelligent partner, just a click away. The assistant can:

• Uncover insights across your software delivery lifecycle.
• Find and trigger self-service actions instantly.
• Guide you in building your Port platform.

Now, you can connect directly to Port AI through an intuitive chat interface where you can ask anything about your software catalog.

Why did we build it?

Back in April, we introduced Port’s AI agents. In the following months, as the agents were in closed beta, we worked closely with design partners and early adopters to capture their feedback. There was a consistent theme - there is gold, but you need to dig for it.

In one customer onsite, we saw that an AI agent they created in Port surfaced insights to critical questions around ownership and hierarchy that they couldn't answer clearly before. 

But, there were downsides:

• Building the AI agent took longer than it should
• Prompts became very technical
• It demanded hands-on assistance from the Port team.

In the meantime, the Port MCP server gained traction and felt like magic. It was very flexible, and it required minimal configuration.

We knew there was potential to replace manual searching and analysis with faster, more intelligent answers.

So what is it?

Port AI assistant answers any question you have about your software delivery process instantly. It can answer questions and perform actions, like:

• Engineer: Which tasks are assigned to me?
• Engineer: Scaffold a new service for my feature
• Engineering manager: What is the status of my team's pull requests?
• SRE: Which services are failing to meet the production readiness scorecard?
• VP engineering: How many P1 incidents did we have yesterday?
• Platform engineer: Help me create a new automation to nudge PR reviewers on Slack

Answering those questions would have taken hours to answer previously.

We’re excited because this release places AI intelligence right inside Port, and we already feel the impact. One team leader at Port cut their weekly status prep from 30 minutes down to 5. An on-call engineer mapped the full blast radius of an incident, catching critical details they normally miss.

‍

‍

‍

What can you do with it?

• Engineer onboarding - New engineers quickly query service ownership and deployment instructions, cutting time-to-first-PR by 30%.
• Automated security - Instantly identify failing production standards and open Jira tickets to fix themi.
• Resource management - Locate idle resources and expensive, non-active environments to optimize budget spending.

How do I start using it?

Using Port AI assistant is simple:

1. Access your Port account (or sign up now for free).
2. Spot the floating button in the bottom-right corner.

3. Ask anything you want.

If you don’t see it, make sure you have access to Port AI features by filling this form. 

How does it work?

‍

The Port AI assistant is an MCP client behind an API. It orchestrates top-tier LLM models and the Port MCP server, using advanced prompts behind the scenes. This API-first approach allows you to query Port AI through an API interface, the user interface, widgets, and Port automations.

To keep your experience safe and secure, we optimized for:

• Permissions - The AI assistant only sees data you can see. It only runs actions you can run.
• Control - actions require human approval, keeping you in the loop of what AI is doing.
• Audit - you can see detailed execution logs of what Port AI did, and which data it accessed. 

What’s next

Port AI Assistant marks a monumental change where intelligence meets you exactly where you work and build. We will continue to evolve this foundation to include AI agents, help you build faster, and surface insights in record time.