Today, we are taking a break from your regularly scheduled technical programming to talk about support. Here at Scout, we consider support one of our differentiators, and even as we adopt AI as a human multiplier behind the scenes, we are committed to keeping it real on the human-interaction side. It will be a long time, if ever, that you reach out to us and get a response from an AI agent. Would it be cheaper? Sure, but it isn’t up to our standards, and we won’t compromise on that. That doesn’t mean we won’t be leveraging internal agents to make our responses faster, more accurate, etc., but it does mean you will have to reexplain yourself multiple times, or give up the quest for an answer altogether when you hit your maximum threshold for AI-chat interactions.
Everyone Does Support
Most companies draw a hard line between support, product, and success teams. In theory, this creates a better experience. In practice, it creates a world where the person who knows your account best isn’t allowed to help you with what you actually need. This holds especially true with technical tools, where interacting with people who don’t understand your use case wastes your time.
We don’t do support tiers or silos of knowledge here; we’re all in the channels. Our head of product hangs out in Discord. Our engineers respond to account and billing questions. Our head of engineering hops on any and all calls. Our head of success and support ships code. This isn’t a mandate. It’s how we make sure we deeply understand what our users actually need so we can keep building the right things.
During a brief stint at a giant SaaS organization that shall remain nameless to protect the innocent, our product manager, Sarah, was once told, “What are they going to do, fire us?” in response to a pitch for better docs and UX bug fixes. We refuse to become that company, and no one on this team wants to work at a place where usability concerns are dismissed as unimportant.
Small on Purpose
We made a deliberate choice to stay focused on like-minded and similarly shaped teams. This gives us luxuries that large enterprises simply can’t maintain at scale, so we acknowledge that this approach won’t be right for every team. We are proud of the support we provide to our larger and more enterprise customers, but when it comes to feature ideation and planning, we don’t skew upmarket.
We’re focused on providing the most valuable application insights to teams like ours: scrappy, curious, get-your-hands-dirty teams who want their app insights when and where they need them and don’t want to spend time troubleshooting their troubleshooting stack when they’d rather be building.
AI Behind the Team, Not in Front of You
You know the drill with most support tools: you type out your problem, and before a human ever sees it, you’re asked to select a category, answer irrelevant questions, attach screenshots, and describe your issue again in a different box. By the time someone actually reads what you wrote, you’ve lost precious minutes.
We skip all of that. A real person reads your request. We use AI on our side to get up to speed fast, pulling relevant docs, previous interactions, searching our codebase, and researching context so we can start the conversation somewhere useful instead of starting it with “can you send a screenshot?”
The Same Tools, Available to You
We shipped MCP servers and a CLI that give AI assistants (Claude, Cursor, whatever you prefer) access to Scout monitoring data. These are the same tools we use internally, and they are available to every Scout customer.
One of our engineers had a production error buried deep in a nested call stack. Couldn’t reproduce it locally. The kind of bug where the debugger doesn’t have anything to latch onto and it’s just you and the source code. Instead of spending an hour stepping through code, he pointed an AI assistant at our MCP server. It pulled the error group, read the stack trace, cross-referenced the codebase, and found the exact line of code in one shot.
That’s not a chatbot deflecting your ticket. That’s AI doing the tedious investigative work so a human can focus on the parts that actually matter.
Your AI assistant can do the same:
- Pull your error groups and start investigating without copying and pasting stack traces around
- See errors in context alongside metrics, traces, N+1 queries, and memory bloat. The full picture of what your app was doing when things went wrong.
- Automate the routine stuff. Summarize new errors, flag regressions, triage before your morning coffee.
- Query from the terminal. The Scout CLI gives you access to app metrics, endpoint performance, traces, error groups, and insights right from the command line. Pipe the output into an LLM and it automatically switches to TOON format, a token-efficient structured format designed for AI consumption.
Whether you prefer the hosted MCP server, the local MCP server running on your own machine, or the CLI piped into your workflow, the data is the same. Pick the path that fits how you work.
Everything is read-only and OAuth 2.0 secured. Setup docs are at www.scoutapm.com/mcp.
The Point
We use AI to make our team faster, not to replace them. When you reach out to us, you get a real person who knows the product because they built it. And we’ve made the same tools available to you so you can work the same way.
Questions? Reach out at support@scoutapm.com.
For application monitoring with errors, logs, and traces, Scout Monitoring provides the fastest insights without the bloat.