Open Source · Apache 2.0

The open-source
alternative to Profound

Track how ChatGPT, Perplexity, and Gemini mention your brand. Self-hosted, with data you own and prompts you write yourself. No demo call required.

TraceAIO vs Profound

Both products track how your brand appears across AI answer engines. How they get there is where they differ.

Feature TraceAIO Profound
Source code Open source, Apache 2.0 on GitHub Closed, proprietary SaaS
Deployment Self-hosted via Docker Compose. Runs on your infra. Hosted SaaS only
Where your data lives Your Postgres database. You own every row. Vendor-managed cloud
LLM coverage ChatGPT, Perplexity, Gemini, Google AI Mode, OpenAI API, Anthropic API 9 surfaces incl. Grok, Copilot, Meta AI, DeepSeek
Real browser sessions Yes, matches what users see Yes
MCP server for Claude 17 built-in tools, conversational queries Yes
REST API & Swagger docs Full API, every endpoint documented ~ Upon request, beta
Webhooks on run completion Configurable URL + Bearer auth Not offered
n8n community node Trigger + action nodes published Not offered
Scheduled runs Hourly, daily, weekly, monthly Yes
Competitor & source tracking Auto-detect, merge, block, citation gaps Yes
Source Watchlist Track specific URLs, sitemap auto-watch ~ Page-level analytics
Time to first run ~60 seconds with one docker compose up Days to weeks. Sales call and POC.
Vendor lock-in None. Export your data or fork the repo. ~ Standard SaaS terms

What you get with TraceAIO

01

You own the data

Every prompt, response, citation, and competitor mention lives in your own Postgres database. You can query it directly or back it up however you want. Your data stays with you.

02

Talk to your data via Claude

The built-in MCP server gives Claude Code and Claude Desktop 17 tools to query your data. Ask in plain English: "which prompts mention our top competitor but never us?" Claude figures it out from your dataset.

03

Real browser, real answers

Browser-based providers run through real browser sessions. What you see is what your customers see when they ask ChatGPT or Perplexity the same question. The answers come from production, not an API approximation.

04

Prompts you actually control

Auto-generate brand-neutral prompts across topics, then add as many of your own as you want. Reuse them across runs and edit them in place. The whole set lives in your config, so it versions like code.

05

Built to plug in

REST API with full Swagger docs, webhooks on run completion, an n8n community node, and an MCP server. Pipe results into Slack, HubSpot, BigQuery, or anything else you already run.

06

Up in 60 seconds

One docker compose up and you're running. The full product, working, on your laptop. Procurement can wait until you've decided you like it.

Try it before you compare

Pull the image and run your first analysis in under a minute. The full product runs on your own machine.

Option A

Docker Compose

Pull the image and start. Everything runs locally, including the browser automation container.

# Download and start curl -O https://raw.githubusercontent.com/jakubsuchy/traceaio/main/docker-compose.yml docker compose up -d # Open http://localhost:3000

Create an admin account, add your API key, and configure your brand in the setup wizard. Switch to Apify in Settings → Credentials if you want faster parallel runs.

Option B

Claude Code

Let Claude set it up for you. Paste this into Claude Code and it handles the rest.

# Tell Claude: Download docker-compose.yml from https://raw.githubusercontent.com/jakubsuchy/traceaio/main/docker-compose.yml and run docker compose up -d Then open http://localhost:3000

After setup, connect Claude to your data via MCP:

claude mcp add --transport http brand-tracker http://localhost:3000/mcp