About

Medoid AI's product lab utilizes our agency's experience to build meaningful AI products. The product lab is currently focusing on solutions for the media and social media sectors. Our solutions give businesses the tools to empower their users' voice, by either removing voice blockers (e.g. hate and toxicity) or by introducing voice enablers (opinion mining and aggregation).

Solution & Product

Hatebusters (HB) addresses the problem of abusive language moderation in user posts entered in online platforms of media companies. Online publishing platforms attract a huge load of abusive posts. This problem is so extended, that some media companies have stopped allowing users to post on their platforms for most of their publications. Our API and software-as-a-service solutions lead to reduced moderation costs and improved moderation quality. Crowdspeak on the other hand, is all about identifying and empowering opinion. In the social media space, individual opinions are effectively muted and lost in the millions of posts and comments every minute. Crowdspeak is our patent-pending user interface element, powered by advanced machine learning, that aggregates and summarizes multiple textual responses given to any single and specific stimulus (e.g. a survey question or a social media post). The aggregation process results in a single short text that expresses the majority of the responses given. People can also participate in a gamified process to be recognized as opinion leaders, alter opinion leaders and other collective opinion formation roles.

Innovation & Scalability

Hatebusters takes an innovative approach on the quality of its AI model by building on the crowdsourcing platform we have developed to collect a significant amount of data, hatebusters.org. Data is the new oil, and an AI model is as good as the data it has been trained on. Hatebusters.org allows us to crowdsource abusive language annotation by volunteers and NGOs while also taking down such abusive content from multiple social networks. At the same time, our model builds upon the state of the art in language understanding technology, that has been pre-trained on huge amounts of text on the web and fine-tuned on hate speech examples in multiple languages. Crowdspeak is a novel and innovative free speech empowerment tool which allows people to shape their collective voice together and researchers to understand public opinion and have access to it for any issue they want to. A unique Crowdspeak ability is to identify the themes and aspects of multiple opinions, and express those in natural language.

Traction & achievements

December 2021 Medoid AI is accepted in the acceleration program of OK!Thess October 2021 Medoid AI wins the patient-centric innovation challenge organized by Pfizer's CDI, Elevate Greece, and the Greek Patients Association, for its solution on plain-language summaries of clinical trials September 2021 Crowdspeak demo is live! https://crowdspeak.medoid.ai May 2020 hatebusters.org in open public beta! (the community platform - please visit!) September 2020 filing a provisional patent in USPTO for CrowdSpeak October 2020 Medoid AI is accepted in the WebSummit startup program for both Hatebusters and Crowdspeak! Their joined pitch passed the 1st round of the pitching competition (out of 4 rounds) December 2020 Internal validation/testing of the Crowdspeak technology - planning public beta for June 2021 January 2021 Hatebusters graduates successfully from the Mediamotor incubation programme (https://mediamotoreurope.eu/)

Team & Why Us

Medoid AI comprises a diverse team of 4 Data Scientists and/or Machine Learning Engineers and a Marketing expert, led by Co-founder and CEO Dr. Anestis Fachantidis, and advised by academic co-founder Prof. Grigorios Tsoumakas. The team has 20 years of experience in Machine Learning and 7 years of experience in NLP. Prof. Tsoumakas is an award-winning academic. Greek news press: - Έλληνες στην κορυφή της επιστήμης, Ιωάννης Π.Α. Ιωαννίδης, 01.10.2018, Καθημερινή - Έρευνα και Ανάπτυξη: Παραλήψεις και λαθεμένα μηνύματα της έκθεσης Πισσαρίδη, Αθανάσιος Παπαϊωάννου, 28.12.2020, tvxs - 6th edition of “Top Scientists Ranking for Computer Science & Electronics” by Guide2Research - Prof. Tsoumakas ranking 27th in the national index for 2020 Media publicity: - "Computers used to teach other computers to play Pac-Man, StarCraft", ArsTechnica, 4/3/2014, Dr.Fachantidis' research collaboration with Washington State University

Funding Needs

Pre-seed investment stage, please communicate with us for more information on a cost-breakdown of our product development plan