Research & Education

ZhuSuan is a Python probabilistic programming library designed for Bayesian deep learning, combining the strengths of Bayesian methods and deep learning. Built upon TensorFlow, it offers a unique approach compared to traditional deep learning libraries that primarily focus on deterministic neural networks and supervised tasks. ZhuSuan provides a suite of deep learning-style primitives and algorithms specifically tailored for constructing probabilistic models and performing Bayesian inference. It supports various inference algorithms including Variational Inference (VI) with programmable posteriors and advanced gradient estimators, Importance Sampling (IS) for model learning and evaluation, Hamiltonian Monte Carlo (HMC) with parallel chains, and Stochastic Gradient Markov Chain Monte Carlo (SGMCMC) methods like SGLD, PSGLD, SGHMC, and SGNHT. The library is still under active development, with installation typically involving cloning the repository and using pip.

Rayyan is an AI-powered platform designed to significantly expedite systematic literature reviews for researchers. It offers AI-powered screening that can reduce review time by up to 90%, alongside a mobile app for on-the-go productivity. Key features include advanced deduplication to eliminate redundant articles, customizable filters, bulk action capabilities, and support for the PICO framework to structure research questions. Rayyan also assists with PRISMA diagram creation, offers a customizable workbench for team collaboration, and provides keyboard shortcuts for enhanced efficiency. The platform supports easy data import from popular tools like Mendeley, Zotero, EndNote, PubMed, and Google Scholar, making it a comprehensive solution for managing and accelerating research workflows.

Mark•r is a revolutionary AI Conversational Assistant browser plugin designed to transform how users interact with web content. By simply highlighting text on any webpage, users can activate a dynamic icon with an AI-powered text field for conversational queries. This tool goes beyond basic highlighting, diving deep into the entire context of the webpage to curate and generate insightful, precise responses tailored to user questions. Mark•r significantly enhances learning and comprehension by providing interactive explanations and customized feedback. It boosts productivity and efficiency by streamlining research, offering quick summaries, and aiding in note-taking. Furthermore, it improves user engagement, assists with content generation and analysis, and provides valuable market research insights, making it an indispensable tool for students, researchers, professionals, and content creators alike.

Kempelen Institute of Intelligent Technologies (KInIT) is an independent research institute dedicated to advancing AI research for both human benefit and industrial application. KInIT's research areas include Web User Data Processing, Green and Secure Environment, Natural Language Processing, and Ethics and Human Values in Technology Operations. The institute offers PhD opportunities, emphasizing hands-on projects and real-world impact within a supportive research community. KInIT also hosts events like the Better_AI_Meetup, fostering discussion on topics such as AI's role in solving ethnicity bias in face recognition. They collaborate with corporate partners and support AI talent development through initiatives like the BrAIn civic association.

Panacea Cooperative Research is at the forefront of applying Artificial Intelligence to forensic science and medical imaging, with a strong focus on human identification. Their offerings include AI-enhanced software for handwriting examination and facial comparison, designed to improve accuracy, efficiency, and objectivity for forensic experts. Beyond product development, Panacea provides services in searching, recovering, and analyzing human remains, offering comprehensive support to institutions and training programs globally. Their research projects span medical imaging and DNA analysis, pushing the boundaries of identification technology. A key product, Skeleton-ID forensic identification software, is utilized by law enforcement and government agencies worldwide for cases where DNA and fingerprints are not viable.

The ARC Prize Foundation is a nonprofit organization dedicated to advancing open-source artificial general intelligence (AGI) research. It achieves this by developing and stewarding scientifically-grounded benchmarks, notably the ARC-AGI series, which measures agentic intelligence and the learning efficiency of AI systems. The foundation also organizes competitions, such as the ARC Prize 2026, offering significant monetary rewards to incentivize novel progress in AGI. The ARC-AGI-3 benchmark, in particular, is designed as an interactive reasoning benchmark to assess an AI Agent’s ability to generalize in novel, unseen environments, moving beyond static evaluations. It provides an ARC-AGI Toolkit and API for researchers to build and test their agents.

The Center for European Research in Trusted Artificial Intelligence (CERTAIN) is a collaborative initiative, legally part of DFKI, dedicated to advancing trustworthy AI. It focuses on developing new technologies that provide functional and other guarantees for AI systems, addressing the often-neglected aspect of trust in international research. CERTAIN researches, develops, deploys, standardizes, and promotes Trusted AI techniques, aiming to provide certification for AI systems. The consortium facilitates project collaboration with both internal and external partners, strengthening European cooperation in responsible, transparent, and human-centric AI.

ChatPDF.ae is an AI-powered tool designed to help users interact with PDF documents effortlessly. It enables users to upload any PDF file and engage in a conversational interface, asking questions, summarizing content, and extracting key insights. The tool provides cited answers directly from the document, streamlining the process of information retrieval. It supports PDFs in any language and can respond to questions in diverse linguistic contexts, making it globally accessible. ChatPDF.ae is particularly useful for students, researchers, and professionals who need to quickly understand complex documents, academic papers, contracts, or manuals, offering a personal assistant experience for document analysis.

The Global Artificial Intelligence Alliance (GAIA) Foundation is dedicated to advancing ethical AI development through a unique collaboration between humans and machines. GAIA conducts scientific research focused on creating empathetic artificial intelligence that supports planetary well-being and human progress. The foundation emphasizes aligning AI with human values to ensure safety and accountability as AI's autonomy increases. Initiatives like Virtual Florence and the Invisible College foster interdisciplinary collaboration and open idea exchange, guiding future actions through a think tank. GAIA also organizes conferences, publishes reports, and engages in advocacy to shape a human-centric, sustainable digital future.

The Institute for Experiential AI at Northeastern University is dedicated to advancing human-centric artificial intelligence. Its core mission involves conducting research that addresses real-world problems and enhances human capabilities through AI. The institute actively collaborates with companies to tackle high-impact challenges, fostering innovation and practical application of AI solutions. Beyond research, it provides valuable experiential education opportunities for both students and executives, ensuring a pipeline of skilled professionals and leaders in the AI field. This dual focus on research and education positions the institute as a key player in shaping the future of AI.

AI Alt Lab is dedicated to building real trust in artificial intelligence by offering actionable alternatives to the current hype-driven, profit-first status quo. Through comprehensive research, education, and evaluation, the organization aims to ensure that AI serves the public good rather than solely private gain. It provides valuable insights and resources for policymakers, founders, and citizens to navigate the complexities of AI development responsibly. The initiative focuses on fostering a deeper understanding of AI's societal impact and promoting ethical, transparent, and beneficial AI systems.

Kashikoi is a comprehensive simulation platform designed for building, evaluating, and rigorously testing AI agents. It enables users to identify and fix bugs in their AI models before they reach customers, ensuring a more robust and reliable product. The platform simulates real-world interactions to autonomously assess agent performance, offering features like AI agent benchmarking with a single prompt and no-code evaluation. Kashikoi supports custom integrations for various AI stacks, allowing users to connect their agents and run realistic simulations against customizable scenarios. It provides actionable insights and synthetic data to optimize prompts, fine-tune models, and enhance overall agent performance, helping users ship better AI agents faster.

Qureight is a tech-first, AI-powered imaging Contract Research Organization (CRO) dedicated to accelerating the impact of new life-saving therapies for lung and heart disease. The platform delivers its suite of products through a proprietary data and imaging curation system, enabling the analysis of clinical imaging and healthcare data. Qureight offers automated image management and AI analysis in therapeutic areas such as Idiopathic Pulmonary Fibrosis, Progressive Pulmonary Fibrosis, Pulmonary Hypertension, and Bronchiectasis. Its offerings include global imaging CRO services, AI-powered endpoints and analytics with proprietary lung image biomarkers (Lung8™, Air8™, Fibr8™, Glass8™, Vascul8™), and data science products leveraging large real-world datasets for synthetic control arms and early drug development support.

The Responsible AI Digest by SoRAI, formerly ABCP, is a specialized publication focusing on the critical intersection of Technology, Society, and Law within the realm of Artificial Intelligence. It provides readers with cutting-edge AI news, covering significant breakthroughs, ethical considerations, and academic research. Building on its legacy, the digest aims to keep professionals and researchers informed about the latest developments and discussions in responsible AI. This resource is ideal for anyone seeking to stay current with the rapidly evolving landscape of AI, its societal impact, and legal implications, offering valuable insights for informed decision-making and research.

The State of AI Report is the most widely read and trusted independent annual report on global progress in artificial intelligence. Published annually since 2018 by Nathan Benaich and Air Street Capital, it aims to spark informed conversation about the state of AI and its future implications. The report covers key dimensions including technology breakthroughs, commercial applications, AI regulation, geopolitics, and safety. It also features a large-scale AI usage survey and predictions for the next 12 months, along with a review of past predictions. All historical editions from 2018 to 2025 are freely available for download.

Student Voice AI offers automated, consistent, accurate, and relevant analysis of student comments for higher education institutions. The platform provides multi-dimensional categorisation, sentence-level analysis, and precise sentiment evaluation, all benchmarked against sector data. It helps universities understand student feedback from various surveys, including NSS, PTES, PRES, UKES, and module evaluations. Key features include intelligent redaction for privacy, comprehensive summarisation, and at-risk student alerts. The tool generates tailored reports for departments and schools, offers sector benchmarking against over 100 UK institutions, and tracks trends over time, enabling data-driven decision-making to enhance the student experience.

WELL-E is the first research and innovation chair of its kind, dedicated to animal welfare and artificial intelligence. This five-year initiative focuses on developing cutting-edge computing innovations to enhance the well-being and extend the longevity of dairy cows across Canada. Funded by the NSERC Alliance program and PROMPT, WELL-E is supported by McGill University, Université du Québec à Montréal (UQÀM), and industry partners including Novalait, Dairy Farmers of Canada (PLC), Dairy Farmers of Ontario (DFO), Les Producteurs de lait du Québec (PLQ), and Lactanet. The project involves various research axes and projects, publications, conferences, and knowledge transfer activities, aiming to bridge the gap between AI and animal welfare.

METR (Model Evaluation & Threat Research) is a research nonprofit dedicated to evaluating frontier AI models. The organization conducts rigorous research and evaluations to improve public understanding of the capabilities and risks posed by advanced AI systems. METR's work includes assessing broad autonomous capabilities, the ability of AI systems to accelerate AI R&D, and potential AI behaviors that threaten evaluation integrity. They also advise AI developers and governments on implementing risk assessment methodologies and publish detailed evaluation reports, sometimes in partnership with leading AI companies like Anthropic and OpenAI. Their research aims to inform policy and promote responsible AI development by scientifically measuring whether and when AI systems might threaten catastrophic harm to society.

TutorOcean is an online tutoring platform designed to connect students with expert tutors and leverage AI technology for enhanced learning. It provides comprehensive academic support for K-12 and higher education across a wide range of subjects, including math, science, languages, and test preparation. Students can search for tutors based on subject, price, and availability, and book sessions directly. The platform features an interactive online classroom with video chat, a digital whiteboard, file sharing, and session recording. TutorOcean also integrates an AI Tutor for quick answers and an AI Copilot for teachers to assist with lesson planning and personalization, combining human expertise with cutting-edge AI.

The Telluride Neuromorphic AI Workshop is a three-week, project-based meeting designed to bridge the gap between neural cognition and artificial intelligence. It focuses on bringing the organizing principles of how brains work into AI development, and conversely, using AI to gain a deeper understanding of brain function. The workshop is structured around specific topic areas, fostering collaborative research and discussion among participants. Key dates include an application deadline, notification of acceptance, and a registration payment due date, culminating in the workshop itself. It also provides access to a LinkedIn community and a mailing list for updates.

AlgorithmWatch is a non-governmental, non-profit organization based in Berlin and Zurich, dedicated to evaluating and shedding light on algorithmic decision-making processes that have social relevance. The organization focuses on how algorithms and AI impact justice, human rights, democracy, and sustainability, aiming to strengthen these values rather than weaken them. AlgorithmWatch conducts research, publishes reports, and engages in advocacy to make algorithmic discrimination visible and promote responsible AI use. Their work includes investigating policy influence, developing guidelines for generative AI, and addressing issues like non-consensual nudity services on platforms. They also offer opportunities for public engagement through data donations and supporting memberships.

Graphzila, powered by GraphQL®, is an open-source query language for APIs and a server-side runtime. It offers a strongly-typed schema to define relationships between data, enhancing API flexibility and predictability. This tool is not tied to a specific database or storage engine, allowing it to integrate with existing code and data, facilitating easier API evolution. It enables users to describe data types, ask for specific information, and receive predictable results. Graphzila is trusted by industry-leading companies and supported by the GraphQL® Foundation, hosted by the non-profit Linux Foundation. It is designed to deliver high-performing user experiences, ensure stability and security for APIs, and promote efficient distributed development.

The Montreal AI Ethics Institute (MAIEI) is an international non-profit organization dedicated to democratizing AI ethics literacy. Founded in 2018, MAIEI equips citizens with the knowledge to understand and address the impact of artificial intelligence on society. The institute publishes a wide range of content including articles on public policy, privacy, human rights, and JEDI (Justice, Equity, Diversity, Inclusion) in AI. It also offers research summaries, an AI Ethics Living Dictionary, and comprehensive 'State of AI Ethics Reports'. MAIEI fosters a learning community and contributes to the discourse on responsible AI through its various columns like 'AI Policy Corner' and 'Tech Futures'.

biometrio.earth offers an AI-powered platform for radical transparency in CO2 offsetting and biodiversity conservation. It measures biodiversity through a holistic approach, combining data from acoustic monitoring, wildlife cameras, and remote sensing with advanced deep-learning models. The platform provides tailored solutions for ecosystem assessment, presenting results via an intuitive Nature Analytics Dashboard. It aims to deliver unparalleled precision and efficiency in biodiversity measurement, supporting organizations in managing nature-based initiatives, ensuring regulatory compliance, and demonstrating impact with confidence. Solutions range from quick insights into biodiversity and land use to comprehensive analytics for entire project areas, suitable for both small and large-scale projects globally.