How an AI agent saved me more than $5K in my way back from Asia

From Thailand to Brazil’s time zone: the playbook I used (with AI) to stop choosing “pretty routes” and start choosing the right route.

If you landed here from my Reels: welcome. Below I share the story, the method and the exact prompt, research flow, toolbox, and my route ranking.

---

Ten tabs open and zero decision

It was late afternoon in Koh Samui. Three tabs: Europe (LIS/IST), Middle East (DXB/DOH), US (MIA/NYC).

Similar prices. Different distances. Housing all over the place. And the annoying question: what path gets me closer to Brazil without breaking the nomad engine that powers my startup?

The unlock was admitting this: I was optimizing only for the ticket. Cheap flight + expensive housing = still expensive.

So many options gave me decision paralysis. Until I decided to eat my own dog food and use AI to fix it.

The turn: total cost > ticket price

I needed an answer that considered everything: flight, housing, cost of living, time zone, work infra, nomad community.

So I activated my favorite web agent (Manus) and turned the problem into a tight brief:

• My profile (airlines, miles, budget, flexibility).

• Decision criteria (productivity > glamour, time-zone drift toward Brazil, total daily cost).

• Clear outputs: a recommended route, ranked alternatives, and a step-by-step plan.

Result? Instead of 48 open tabs, I got a structured plan: 1 primary recommendation + 2 solid “Plan B” routes plus clear next steps and almost $ 5.000 in savings. Unreal.

Three ideas that changed the game

1. Build the research prompt itself using AI (in this case I've used ChatGPT 5 Thinking)

2. Stepping stones: hop through hubs that shift the time zone and lower stay costs before crossing the Atlantic.

3. Use Total Cost of Travel to optmize: compare route bundles (flight + 7-day stay + productivity) instead of isolated tickets. The “cheap flight” is expensive when the rest of the system doesn’t pencil out.

---

What you take from this (free):

1. Route math must be holistic (ticket + housing + time zone + work).

2. Use AI as a research operator, not just a flight finder.

3. Ask to return in “stepping stones” destinations to maximize cost–benefit and energy

🔒 For subscribers: my complete playbook (prompt, flow & step-by-step guide)

Oh Iago, really? Yes, it's less than $5 per month and you get access to this post but also all my other ones :)

Enjoy, it's less than a coffee per month! And by the way you are probably a friend and know that I'm serious.

Included: my exact Manus prompt that you can use in any tool you prefer, a video explaining how to use, the reverse engineered process I used, and my real example with 3 best cost–benefit routes for leaving Asia and moving closer to Brazil’s time zone.

Video showing the exact step by step process I used so you don't have to :)