The Next Generation of Kngine beat Siri, Wolfram | Alpha and Evi by 3X Answering 400 Million Questions
Kngine is Question Answering engine similar to IBM Watson and Wolfram | Alpha (which power Siri). released the iOS and Android App while ago. Kngine is result of 4.5 years of research + over 1M line of code.
IBM utilize statistical model, and it run on supercomputer yet it’s slow (it answer only one question at a time and each question takes about 4sec.). Unlike IBM Watson Kngine utilize Knowledge-Based technology. Kngine continuously read the Web and it tries to understand its contents and build a memory of the knowledge, so when you ask it questions such as ‘What is the average weight of a lion’ it will understand it as you do and bring you the exact answer not links from his memory. Even when you ask him about something it doesn’t know about yet, it will try to find the answer for you and will add this topic to its learning queue.
Kngine answered over 400M questions till this moment and the Public version (v 1.0) of Kngine scored 2X compared to Siri (Wolfram | Alpha) and Evi in the 3rd party benchmark that was conducted. The next version Kngin plans to deploy after it close the current funding round scored 3X compared to any technology in the market (71%, compared to 25% for Siri (Wolfram | Alpha) and Evi. Kngin goal in 2013 is to beat IBM Watson while consuming 10% of IBM Watson computation power, and eventually it will play Jeopardy versus IBM Watson.
The startup was founded in Cairo, Egypt got fund of 275K from Sawari Ventures. And now HQ is in Silicon Valley and the Engineering office in Cairo, Egypt.
Kngin plans to open Question Answering APIs that can power any application, and also plans to integrate Kngine QA inside out applications as well (API’s and tech license)
Impressive Global startup coming from Egypt we hope we saw more of those.