While attending Structure in San Francisco recently I was briefed by import.io about there solution. During the event I probably did a dozen briefings but import.io really stood out to me. What they’re doing is in really fact pretty simple but, like all good solutions, the simplicity at the front end hides a high degree of complexity at the back end.
So what is import.io? Essentially import.io is a data tool that allows users to consider the entire web as one big database. We all implicitly know that the web is the biggest source of data on the planet, but have always struggled to find manageable ways to enable the harnessing of that data. Customizations, specificities and nuances – all the things that make the web the highly flexible place that it is – also make it exceptionally hard to find some common structure across multiple data sources o the web.
This is where import.io comes in – The service allows users to harness the data by removing the complexity that differences of data structure create. Data can be extracted from multiple websites, normalized and combined to create a single dynamic database which can then be manipulated to create visualizations and other outputs.
Traditionally this sort of process required specific screen-scraping technologies to be built – the problem with this is that it requires developer time, is specific from site to site, and breaks every time website structure changes. With import.io data acquisition from the web is business analyst rather than a developer role. The company has several distinct offerings:
- Data Browser (io) – a browser with powerful additional features that allow users to transform the sites they visit into an actionable dataset
- Cloud Data Platform – the cloud backend powers the federation and extraction, making it accessible via the io browser or the io API
- Universal Access & Federation – Tools that ensure data published on the web is accessible – dynamic pages, forms & APIs are all easily consumed
In my briefing I was shown how import.io can aggregate data from a number of different real estate websites, normalize that data and populate it into a common database – but the potential user cases are many. According to the company, Import.io has already gained traction with companies such as international recruiting company Robert Half, Bloomberg and Hewlett Packard who all use import.io to aggregate data from the web and present it to users in an actionable format.
There is a certain poetry to a solution that, unlike the hand waving and rhetoric around the death of structured data, takes a mass of data from the web and delivers it up in a traditional columnar format. I really like what import.io is doing and am certainly looking forward to seeing how they develop the commercial applications for their product.