/
Anywhere
/
Anywhere

For property, recruitment and marketplaces

Anywhere

Ask a location question in plain English. A language model plans the calls, and TravelTime’s APIs answer it on a map.

How it's built

  1. Puts a pin on every place the question namesGeocoding
  2. Maps everywhere each commute reaches on a typical weekday morning, and keeps the overlapH3
  3. Keeps the areas near what the question asks for, and checks noise, crime, flooding, broadband and who lives thereData API
  4. Ranks the neighbourhoods left and counts the people in themData API
  5. Routes the real journey from the best areas and re-ranks themRoutes

Build it yourself

The same calls from your own code, with the official SDK. A free key gets you started.

JavaScript

import { TravelTimeClient } from 'traveltime-api'

const client = new TravelTimeClient({ applicationId, apiKey })

// Everywhere a commute into each workplace reaches on a weekday morning
const { results } = await client.h3Fast({ resolution: 9, properties: ['mean'], arrival_searches })

Where it fits

Property portals, job boards and marketplaces add natural language search on their own backend: a language model plans the calls, and TravelTime supplies the journeys and the neighbourhood data.

Which TravelTime APIs does Anywhere use?

Geocoding for the places a question names, H3 for where each commute reaches, the Data API for what is nearby, the neighbourhood and who lives there, and Routes for the real journey from the best areas.

What does the language model do?

It plans: it reads the question, picks the calls and writes the answer from what they return. Every time, place and count comes from TravelTime, and How it works shows each request.

Do I need an expensive model to build this?

No. The model only chooses calls from a short list of tools, so any model that can call tools fits. This demo uses Claude Sonnet 5.5, at a few cents a question.

More demo apps

Build something like this

Tell us what you are building, and we will show you which APIs fit.