Product Design

Curiosio
5 min readOct 16, 2021

by Vas Mylko

This post is about the further design of Curiosio — the smartest and coolest road trip planner for the new world. There are a few versions ahead of us until the level of smartness and coolness is what we and travelers want.

User Voice

Since the public release of Destination Trips aka Beta15 on Oct 1st, we gathered user feedback. Some reports and requests were the same as before, some were similar, and some were new — unlocked opportunities by the latest version of Curiosio. Below is a list of commonalities that travelers want in comparison to what Curiosio can do today:

  • Enter any POI or place by its name, even by alternative names (e.g. Lake Braies vs. Pragser Wildsee vs. Lago di Braies). So far Curiosio allows entering the points that are cities, towns, some villages, parks.
  • Enter a region by name (e.g. Napa Valley, Val d’Orcia). The region could be bigger than a city/town/park. The region could be a part of the city/park.
  • Visual inspiration on the Landing Page. Travelers want visuals, entertainment. Travelers don’t want simplistic Google-style pages because they expect excitement from Curiosio.
  • Visual preview of recommended POIs/places without explicit navigation to Wikipedia or Atlas Obscura. Must be in-place preview, ad hoc in a pop-up.
  • Paste POI/place names one by one into the list of travel-through points. So far Curiosio accepts bulk-paste in special Trip Markup Language that allows pasting the entire search context as semi-plain text at once.
  • Interaction with the initial trip plans created in the search context. By seeing what’s possible the traveler starts to understand what she likes and what she wants. Features to refine the style of the trip are needed (e.g. More Like This, Less Like That).
  • Fine-tuning of the trip plan that almost fits the traveler’s needs and requirements. The traveler is OK to do the necessary edits herself.
  • Implicit and explicit interaction with the underlying AI to make it work more personalized. Explicit is via fine-tuning the configuration, by turning some knobs and pushing some toggles.
  • Enter a custom point by geo-coordinates, and custom name it. The context is known only to the traveler who is making the custom point.

Same Scenes Different Place

Travelers want to see the world their own way. But how exactly? What is your own way? There are different interests, there are different intentions. Everybody is seeing and experiencing the world through a personal prism. The visual problem with our Landing Page could be solved with some kind of prism enabling the personal look at the world. Let me introduce The Prism concept.

Travel geek with The Prism

Some travelers pay attention to historical heritage, architecture, style, culture, natural scenery. Other travelers like to meet people, feel the vibe, try the food, fashion, nightclubs, games. We are going to redesign the Landing Page to show countries via different prisms. Several facets-scenes per country could be seen simultaneously at a time. You will be seeing the same scenes in different countries.

Depending on what prism was active when the traveler opened the country (or region) page Curiosio will be preparing the high-level theme of the trip there. Pilot prisms will be Curious, Obscure, and maybe Serendipitous. Need a shorter word for the last one, any ideas?

Wireframe for The Prism on the Landing Page

We will think about porting The Prism to arbitrary geographic regions.

All Scenes Same Place

Curiosio could show you more facets-scenes of the current prism for the country on the Country Page. E.g. on the Landing Page, you saw The Prism with four scenes for France, Italy, United States — all via Curious Prism. You selected France and got navigated to France Page. Now you could see a dozen of scenes in France, via the same Curious Prism. You will be seeing all scenes in the same country.

Wireframe for The Prism on the Country Page

There are going to be a dozen of scenes on the Country Page. Those facets-scenes will be interactive. You will be able to like/dislike a scene by scene to define the detailed theme. We are thinking about a scrollable list of scenes where you could swipe or click “More like this” or “Less like that” in a Tinder-like manner.

Human-AI-Human

Curiosio is AI-powered. AI engine Ingeenee is working over the Travel Knowledge Graph. The interaction session is a “hamburger” of Human, AI, and Human again. Human above AI, Human below AI.

Human above AI

Human interaction before AI defines the goal and mission. A typical use case that we call a “Social Soup” is this:

  • A traveler is getting inspired on Instagram, YouTube, Google Maps, Wiki, Facebook.
  • The traveler gathers the points, sees the perimeter, finalizes the travel-through list, selects a point to stay at.
  • The traveler enters the desired travel-through points into Curiosio, sets the staying point as a destination, sets duration and budget, and now wants the daily plans to visit everything properly.
  • The traveler is interested in what else is possible in that duration and budget in the vicinity of the destination. Technically/academically speaking this is a math problem of draconian complexity. Astronomical complexity.

AI between Humans

Then, AI takes over. Taking this problem statement the AI performs the data crunching and magic and produces several trip plans as the search result. Those trip plans may satisfy all requirements in full, or some requirements may be not met because there is no solution [in the computational universe] for that context.

Our AI engine Ingeenee will take the Travel Knowledge Graph and transform it according to The Prism on the Landing Page into the Interest Graph. Then it will transform the Interest Graph into Intention Graph according to The Prism on the Country Page, from your interaction with it. Then AI will solve the NP-Complete math problem to make your interesting trip plans in your personal requirements in predictable real-time.

Human below AI

It’s the same Human who was above the AI and will be below the AI — it’s our traveler. After seeing the possible trip plans and other trip plans that satisfy slightly augmented requirements the traveler is entering the stage again — she wants to customize because of gained new information from the search results, and by feeling emotions during the session.

Here nothing suites the personal needs better than Do It Yourself. Final fine-tuning of the trip plan is done by the traveler herself. Fine-tuning is quite a scope of work, it deserves a separate post. Stay tuned for the preliminary wire-frames and ideas.

Meanwhile, tweak Curiosio to make it work for you — many workarounds are possible even with the current limited functionality. And always follow your curiosity!

--

--