< Company Intro >

Overview:

Established at the end of 2015, we are a Techstars company with 35 employees in Korea with a subsidiary in London as well.

Our CEO is a search mafia in Korea, who has worked in the search industry for more than 20 years. He is the first person to commercialize semantic search in Korea.

What we do:

We have built a QA(Question Answering) system based on Deep Learning, which is different to the existing players who heavily rely on the rule-base or the ontology system.

Our core strength is that we can understand the true intention of a question and can find the related passage using Deep Learning and MRC (machine reading comprehension).

As a public reference, we have co-ranked #1 in SQuAD 2.0 with Google AI team, which is an MRC competition led by Stanford University’s NLP team.

Use cases:

Our patented QA system is the 1st and only commercial version in Korea which is already used in 50% of the smart speakers in Korea. Since our launch, we have worked on projects including a Smart Watch for LG, a Customer Service Chatbot of LG, a Smart Speaker of Korea Telecom, a Data Intelligence System for Daewoo and a Korean integration for Softbank’s Pepper robot.

SaaS related:

We will be providing our system in a SaaS platform this year. 3 major types of our PaaS platform deployment would be 1) On-premise, 2) 3rd party (API, SDK) and 3) Open API.

Funding & Investment related:

We have completed our seed funding including gov’t grants of $3.6m. Our seed funding was led by NAVER, which is the most dominant search engine in Korea, with a market cap of $20bn.

We are now seeking for a Series A round which will be up to $10m

We are currently talking with investors in Korea, Europe, and US.

< The Short Story of 42Maru >

42Maru is building the next generation of search. Our QA (Question Answering) system uses deep learning to understand a question and deliver a single correct answer from a large unstructured data-set. Our technology finds the answer in real time using MRC (Machine Reading Comprehension). We have co-ranked #1 with Google AI team in Stanford University’s MRC competition. (as of Nov. 2018)