Maluuba

History

Maluuba was founded by two undergraduate students from the University of Waterloo, Sam Pasupalak and Kaheer Suleman. Their initial proof of concept was a voice activated travel tool that allowed users to search for flights using their voice.

In February 2012, the company secured $2 million in seed funding from Samsung Ventures. Within the span of 6 months following the investment, the company built out an engineering team, technology platform and Android personal assistant application that rivalled Siri and Google Now in terms of functionality. In September of that year, Maluuba officially launched their Android application as a finalist on stage at the Techcrunch Disrupt.

Having noticed the launch and early market success of personal assistants like Siri and Google Now, consumer electronics companies and other device makers became interested in incorporating the technology into their products. Since 2013, Maluuba has partnered with several companies in the smart phone, smart TV, automotive and IoT space. For instance, Maluuba's personal assistant technology powers LG's VoiceMate application found on the company's flagship G series smartphones.

Maluuba's vision from the beginning has been to bring human level literacy to machines. In August 2015 Maluuba secured a $9 million of Series A investment from Nautilus Ventures and Emerllion Capital. Then in December 2015, Maluuba opened an R&D lab in Montreal, Quebec, widely considered one of the epicentres for deep learning research.

The company has grown to more than fifty employees and in 2016 published fifteen peer-reviewed research papers focused on language understanding.

On January 13, 2017, Maluuba announced they have been acquired by Microsoft.

Research

Maluuba's deep and reinforcement learning research centre opened in Montreal, Quebec in December 2015. The lab is advised by Yoshua Bengio (University of Montreal) and Richard Sutton (University of Alberta). Recognizing the importance of its research to the broader AI community, the lab has published fifteen peer-reviewed papers discussing some of its recent research. The lab also partners with the University of Montreal MILA lab and McGill University.

In late 2016 the company released two natural language datasets, NewsQA focused on comprehension and Frames focused on Dialogue.

Machine Reading Comprehension

The research lab works on machine reading comprehension, publishing several papers and results using well known machine reading benchmarks like MCTest. Maluuba has achieved 80% accuracy on the MCTest, outperforming other word-matching approaches by 8%, and surpassed the previous benchmark set for deep learning techniques, the DSTC2, by 3%, bringing it to 83%.

EpiReader

In June 2016, the company demonstrated a program called EpiReader which achieved state of the art results at the time and outperformed Facebook and Google in machine comprehension tests. Several research teams were able to match Maluuba’s results since the paper was released. EpiReader made use of two large datasets, the CNN/Daily Mail dataset released by Google DeepMind, comprising over 300,000 news articles; and the Children’s Book Test, posted by Facebook Research, made up of 98 children’s books open sourced under Project Gutenberg. According to Yoshua Bengio, a pioneer in the field of deep learning " [Machine reading comprehension] technology will be in every kind of user interface in the future.”

Dialogue Systems

The company has published research findings into dialogue systems which comprises natural language understanding, state tracking, and natural language generation. Maluuba drew insights from human behaviour and published a research paper on policy manager (decision maker) where the system is rewarded for a correct decision. In 2016, Maluuba also released Frames dataset which is a large human-generated corpus of conversations and was made available freely to the AI research community. Maluuba aims to improve the conversational capabilities of chatbots and virtual assistants by applying deep learning techniques.

Reinforcement Learning

The company conducts research into reinforcement learning in which intelligent agents are motivated to take actions within a set environment in order to maximize a reward. Maluuba’s researchers are laying a foundation towards artificial general intelligence by applying RL techniques to a single algorithm that can work on images, games, and text. The research team also focuses on scalability and has published several papers.

Applications

Numerous applications for Maluuba's technology have been proposed in industry with several applications being commercialized.

One of the first applications of Maluuba's natural language technology has been the smartphone assistant. These systems allow users to speak to their phone and get direct results to their question (instead of merely seeing a sea of blue web links that point to possible answers to their question). The company raised $9M in 2015 to bring their voice assistant technology to automotive and IOT sectors.

Maluuba also offers a Google Chrome extension, NewsQA, that uses deep learning algorithms to read through news articles in order to answer questions posed by the user.