§ AI Chat
Fifty-five research tools behind one search box.
Ask in plain English. A tri-stage pipeline - regex, fast model, deep model - routes every question to the right tools, then searches sailings, scores value, and compares lines and ships. About $0.003 per query. Ten queries included in the trial.
Start 14-day trial49 lines · 678 ships · 68,722 sailings searchable
cruisingintelligence.com / app / ai-chat
How the pipeline works.
Behind the search box sit 55 research tools covering sailings, ships, cruise lines, ports, and value scoring. You never pick which one to call - the pipeline routes the question for you.
Stage one is pure regex: direct lookups like a ship name or a region resolve against the index instantly, with no model involved. Stage two is a fast model that parses structured questions - filters, sorts, availability - into tool calls across the sailing database. Stage three is a deep model, escalated only when the question needs multi-step reasoning: comparing lines, weighing trade-offs, scoring value across a shortlist.
Most questions never get past stage two. That is how a query that touches dozens of tools stays around $0.003 - and why ten queries come included in every trial.
What it answers.
Sailing search
Every departure across the fleet, searched by destination, date, duration, budget, or ship - in one pass.
Value scoring
Sailings scored 0-100 on what you actually get for the price, not just the sticker rate.
Line & ship comparison
Side-by-side comparisons across cruise lines and ships, weighted by what matters for the client.
Cost per query
About $0.003 per query. Ten queries included in the 14-day trial.
Example answer
"3 lines match across 41 sailings. Princess leads on value (82/100) with 14 departures from $212/day."