What Hugo does
Hugo is an autonomous AI agent built on top of RoastIQ and BuyerLens. Instead of navigating between tabs and configuring studies manually, you talk to Hugo. He decides which tools to use, runs them, reads the results, and delivers the answer.
How it works
You type a message. Hugo reads it and decides what to do:
- If you uploaded a video, Hugo runs RoastIQ automatically
- If you ask about a weak KPI, Hugo runs BuyerLens interviews
- If you ask for a comparison, Hugo pulls both reports side by side
- If you ask for a brief, Hugo synthesizes everything into an edit direction
Hugo has 8 tools: run RoastIQ, read reports, run BuyerLens, search benchmarks, compare creatives, generate briefs, check projects, and query the database.
What makes Hugo different from a chatbot
A chatbot answers from training data. Hugo reads your actual data, meaning your RoastIQ scores, your benchmark context, and your buyer interview results, then reasons about what to do next.
When Hugo says move the product demo to seconds 3 to 5, that recommendation traces back to your specific diagnostic rather than to a generic prompt. You can check it against the evidence timeline that produced it.
The ordering rule still applies
Hugo does not bypass the pipeline. BuyerLens still requires a RoastIQ result first, whether you run it yourself or ask Hugo to run it. If you ask Hugo about buyer resistance on a creative that has not been scored, he scores it first, because an interview without diagnostic context produces generic output.
Thinking states
Hugo shows you what he is doing: searching benchmark data, reading the RoastIQ report, deliberating on the evidence, composing the recommendation. You see the reasoning, not just the answer.
Try Hugo
Hugo is available in the workspace at /app/hugo. You can also learn more on the Hugo landing page at saliencylab.com/hugo.