Diffbot, a super cool visual learning tool, or robot, that can see the web just like humans do, has just raised $2 million in seed funding from a collection of tech investors, including Joi Ito, Director of MIT Media Lab; Elad Gil, VP of Corporate Strategy at Twitter; Jonathan Heiliger, former VP of Technical Operations at Facebook; Redbeacon co-founder Aaron Lee; and founder of VitalSigns Montgomery Kersten.  CEO Michael Tung plans to use the new funding for new hires and to to expand its resources.  

Diffbot’s robot technology is an API, which lets people navigate the hidden world of the Web visually. To use Diffbot, a developer or application submits an URL link and Diffbot will return structured data that a person can easily understand the different sections of a website, such as the title, author, text, image and video. Diffbot also knows the context of the content, so that Apple, the computer brand, is clearly different than Apple, the fruit.

Diffbot launched several sets of APIs last summer, and now the startup currently process 100 million API calls per month from companies seeking to organise web content. Its customer base includes thousands of developers who are using the APIs but its customer paying base is much smaller.  However they do have AOL, StumbleUpon and Hacker News Radio as part of their success stories.

Currently, Diffbot has categorized the web into about 20 different page types, but its main usage now are homepages and article pages.  In the near future, Diffbot will be able to recognize other types of pages, such as product pages, review pages, recipe pages and even social networking profiles.

CEO Mike Tung started Diffbot as a graduate student at Stanford.  While taking 8 CS classes in one quarter, he created a program to monitor all his classes webpages. When his friends wondered how he was always staying informed and in real-time too and asked if they could use it, he realized, his bot technology could be used not just for his class webpages bit for any page on the web.

Even though Diffbot is meant for developers and applications, you can use Diffbot for free.  Just go to their website and see what their API can do.  I just did it by copying and pasting a blog link from TigerStartups and I have to say it returned and labeled the content accurately.  Try it at the least because it’s just cool!