Research & Education

The Max Planck Society is Germany's leading non-university basic research organization, operating 84 institutes dedicated to advancing science. It covers a vast array of research areas, including artificial intelligence, climate and sustainability, biodiversity, genome editing, and synthetic biology. The society emphasizes interdisciplinary collaboration and offers numerous programs for early-career researchers, including International Max Planck Research Schools and postdoctoral programs. It also focuses on technology transfer and entrepreneurship, fostering innovation through patents, licenses, and start-ups. The Max Planck Society is committed to addressing societal challenges through its research and promoting scientific discourse.

HydroX AI provides an end-to-end platform for secure and responsible AI adoption, focusing on safety, compliance, and red-teaming solutions. It empowers organizations to protect their AI models, data, and users in an evolving AI landscape. Key features include continuous AI safety evaluation to identify risks before and after deployment, real-time AI monitoring to reinforce security and privacy across live model outputs, and customizable AI risk assessments tailored to industry, domain, or use case. The platform helps organizations deploy AI systems safely, responsibly, and in compliance with evolving AI regulations, ensuring trusted and compliant AI.

Scira AI is an advanced AI assistant designed to streamline research and information gathering. It offers in-depth web searches, provides citations for its findings, and integrates with over 100 applications such as GitHub, Notion, and Slack. Users can leverage its capabilities for extreme deep research, PDF analysis, and various search modes covering web, academic, social media, and specialized topics like crypto and stocks. The platform supports multiple AI models, including Claude Haiku, Sonnet, Opus, and Gemini 3.1 Pro, catering to diverse research needs. Scira AI is ideal for students, researchers, and professionals seeking to enhance their learning and productivity.

Citely is an AI-powered platform designed to enhance research credibility by offering robust citation verification and source finding capabilities. It allows users to check fake citations, verify references, and locate original sources across a wide array of academic databases including CrossRef, PubMed, arXiv, and Google Scholar. The tool features an AI Source Finder for instant source identification, a Citation Checker Online to detect fabricated references, and an AI Research Assistant for real-time guidance on citation authenticity. Citely supports all major citation styles like APA, MLA, Chicago, Harvard, and IEEE, and integrates with popular reference management software such as Zotero, Mendeley, and EndNote. It aims to streamline the research workflow and maintain high academic standards.

cnn-watermark-removal is an open-source project providing a fully convolutional deep neural network designed to remove transparent overlays, such as watermarks, from images. The architecture avoids down-sampling to maintain accuracy and utilizes dilated convolutions for a large receptive field without excessive computation. While the current design does not generalize well to vastly different watermark types or non-additive overlays, it offers a solid foundation for image inpainting. Users can train the network with their own datasets or use provided pretrained weights. The project is built with Tensorflow and requires specific libraries and datasets for setup and operation.

General Trajectory is focused on advancing general-purpose AI to solve real-world problems within the physical world. The company, part of Y Combinator's W25 batch, aims to leverage artificial intelligence to foster an abundance of goods and services. While specific features are not detailed on their current website, their core mission revolves around applying AI to tangible, physical applications rather than purely digital ones. This suggests a focus on areas like robotics, automation, and other physical infrastructure where AI can drive significant improvements and efficiencies.

Himitsu Lab Limited is an advanced AI research organization established in 2019, dedicated to exploring intelligent systems, reasoning under uncertainty, and safe deployment of next-generation artificial intelligence. Their foundational research focuses on Native AI Intelligence, aiming for internal world understanding, uncertainty-aware reasoning, decision realism, and safe real-world action across various environments. Key research areas include World Model+ Decision Architecture, Cognitive & Reasoning Models, Swarm & Distributed Intelligence, Quantum Security & Long-Horizon Risk, and Physical & Embodied AI. The World Model+ framework is central to their approach, extending classical world models with features like uncertainty-aware reasoning, decision timing, and continuous learning to ensure safe operation under real-world constraints.

The Precedent AI offers a regulatory intelligence engine designed to embed legal understanding directly into a company's products and processes. It helps legal teams identify relevant bills, regulations, and case law, and understand the impact of changes. The platform processes millions of regulatory updates to pinpoint those critical to specific products and subsidiaries, avoiding generic 'industry updates.' It integrates with existing workflows, pushing actionable intelligence to channels like Slack, Teams, or Email, and uses AI agents to autonomously scan and map regulatory developments to company context, ensuring research is complete before the workday begins.

MaxisIT Inc. provides comprehensive AI-driven clinical data management and analytics solutions specifically designed for life sciences organizations. Their platform, Maxis AI, automates data management and analytics processes, ensuring higher-quality data and faster insights for clinical trials. Key products include the Clinical Trials Oversight System for real-time visibility, Data Management Workbench for seamless integrations, and Statistical Computing Environment for accelerated research. The platform also features a Clinical Data Repository, Risk-based Quality Management, Smart Optimizer for AI-enabled analytics, and a Metadata Repository. MaxisIT aims to accelerate drug development timelines and revolutionize clinical data management by offering a single source of truth for clinical study information, supporting roles from clinical operations to executive management.

PatentWatch is an AI-powered platform designed to streamline patent intelligence and monetization. It enables users to efficiently map patents to products at scale, conduct comprehensive patent searches, and perform in-depth prior art analysis. The tool generates detailed claim charts, providing AI analysis scores for each claim element against evidence, helping to identify potential infringements and assess patent strength. PatentWatch aims to accelerate IP work, saving significant hours per claim chart and leveraging a vast patent database to help users turn their intellectual property into revenue.

Sravathi AI Technology Pvt Ltd pioneers AI-driven solutions for complex drug discovery and development challenges, founded in 2020. The platform leverages generative AI, predictive AI, and physics-based models to design novel compounds, identify and validate targets. It allows clients to rapidly narrow down to a few hundred prioritized compounds for chemical synthesis by considering physico-chemical, ADMET, and pharmaco-kinetic properties. This approach significantly reduces uncertainty, time, and costs in the drug discovery process, transforming chemistry with its cost-effective, rule-based, and data-driven AI platform.

Melwy specializes in point-in-time (PiT) LLMs designed to eliminate look-ahead bias in financial models. This innovative approach allows for more realistic backtesting, advanced agentic trading research, and reliable deployment of models beyond their initial training window. The platform offers PiT-Inference models that demonstrate significantly reduced alpha decay compared to standard LLMs, with some models boasting over 500 billion parameters. Melwy is geared towards professionals in finance who require robust and generalizable AI solutions for trading and research, providing a critical advantage in a data-driven market.

DeepResearch, developed by Alibaba-NLP, is an open-source agentic large language model specifically engineered for long-horizon, deep information-seeking tasks. With 30.5 billion total parameters and 3.3 billion activated per token, it demonstrates state-of-the-art performance across various agentic search benchmarks. Key features include a fully automated synthetic data generation pipeline, large-scale continual pre-training on agentic data, and end-to-end reinforcement learning. It supports both ReAct and IterResearch-based 'Heavy' inference paradigms, allowing for rigorous evaluation of intrinsic abilities and maximum performance. The model is available for download and can be integrated via OpenRouter for inference without GPUs.

DeepSeek-MoE is an innovative Mixture-of-Experts (MoE) language model featuring 16.4 billion parameters. It utilizes a unique architecture with fine-grained expert segmentation and shared experts isolation, allowing it to achieve performance comparable to DeepSeek 7B and LLaMA2 7B while requiring only about 40% of the computations. Trained from scratch on 2 trillion English and Chinese tokens, DeepSeek-MoE provides both base and chat model checkpoints for research and commercial use. It can be deployed on a single GPU with 40GB of memory without quantization, and offers quick start guides for installation and inference using Huggingface's Transformers. The project also provides scripts for fine-tuning the models on downstream tasks, supporting both DeepSpeed and QLoRA configurations.

The History Timeline Generator is an online AI-powered tool designed to create customized timelines for historical events. It assists users in organizing and presenting the development of historical occurrences, making complex information accessible and understandable. The tool allows for the generation of timelines for various subjects, including historical figures, fictional narratives, country histories, and specific scientific or cultural developments. Users can search for public cached timelines instantly or generate new ones, with queries consumed only when AI is needed for new or regenerated content. This makes it an invaluable resource for students, educators, and history enthusiasts looking to visualize and explore historical data efficiently.

GREYC is a prominent research laboratory, a joint unit of CNRS, ENSICAEN, and UNICAEN, dedicated to advancements in digital science. Their extensive work encompasses critical areas such as informatics, image processing, and instrumentation. The laboratory boasts a significant team of over 180 members and has contributed to more than 2100 publications over 25 years. GREYC actively engages in various research axes, teams, seminars, and transversal groups, fostering a dynamic environment for scientific exploration and innovation. They also highlight industrial collaborations and offer opportunities for employment, including permanent positions, theses, and internships.

OSWorld is an open-source benchmark designed to evaluate multimodal AI agents performing open-ended tasks in real computer environments. It offers a robust framework for researchers and developers to test and compare the capabilities of their AI agents. The platform supports various virtualization technologies like VMware, VirtualBox, and Docker, with ongoing support for cloud platforms such as AWS. Key features include parallel execution of experiments, detailed result logging with screenshots and video recordings, and tools for manual task examination. OSWorld aims to standardize the benchmarking process for AI agents, providing clear metrics for success rates across different domains like Office, Daily, and Professional tasks.

Paper2Agent is an innovative multi-agent AI system designed to automatically convert research papers into interactive AI agents. This system requires minimal human input, streamlining the process of transforming complex research into actionable AI tools. It intelligently identifies and executes relevant tutorials directly from a research paper's associated codebase, enabling users to interact with the paper's methodologies and findings in a dynamic way. Paper2Agent supports both basic and advanced usage, allowing for targeted tutorial processing and integration with API keys for authenticated repositories. The system generates a comprehensive project structure including an isolated Python environment, extracted tools, and detailed reports on code coverage and quality, making it a robust solution for researchers and developers.

PhiFlow is an open-source simulation toolkit designed for machine learning and optimization, primarily written in Python. It offers a differentiable PDE solving framework that seamlessly integrates with popular machine learning frameworks such as NumPy, PyTorch, Jax, and TensorFlow. This integration allows users to leverage automatic differentiation for building end-to-end differentiable functions that combine learning models with physics simulations. PhiFlow supports a wide range of applications, particularly in fluid dynamics, with features like built-in PDE operations, a flexible web interface for live visualizations, and object-oriented design for extensibility. It enables reusable simulation code across different backends and dimensionalities, making it a versatile tool for researchers and developers.

PyTorch-StudioGAN is a Pytorch library designed for implementing and researching Generative Adversarial Networks (GANs) for both conditional and unconditional image generation. It serves as a unified playground for machine learning researchers to compare and analyze new GAN ideas, offering implementations of 7 GAN architectures, 9 conditioning methods, 4 adversarial losses, and various regularization and augmentation modules. The library also provides an unprecedented-scale benchmark for generative models, including results from GANs, auto-regressive models, and Diffusion models. It supports multiple acceleration methods like distributed data-parallel training and mixed-precision training, ensuring flexibility and reproducibility in GAN research.

Bagoodex provides an online AI chat experience, functioning as both an AI chatbot and an AI chat generator. Users can leverage the platform to search for information, generate written content, and create images instantly. The tool emphasizes a fast, flexible, and intelligent interaction, aiming to enhance information retrieval and content creation through its AI capabilities. The website indicates a recent update, including a new domain and refreshed features, with a prompt to "Move to Sigma AI," suggesting a rebranding or migration.

Local Deep Research is an AI-powered research assistant designed for deep, agentic research, leveraging multiple LLMs and search engines. It supports both local and cloud LLMs, including Ollama, Google, and Anthropic, and integrates with over 10 sources like arXiv, PubMed, and the web, alongside private documents. The tool emphasizes privacy, allowing users to run it locally with encrypted data and no telemetry. It achieves high accuracy on benchmarks and offers features like quick summaries, detailed research reports, document analysis, and an autonomous LangGraph Agent Strategy for adaptive searching and synthesis. Users can build a searchable knowledge base from downloaded sources, ensuring data ownership and security.

The MLIA team, part of the LIP6 laboratory at Sorbonne University, specializes in Statistical Machine Learning with a strong focus on algorithmic development and applications in semantic data analysis. Their work involves international and national collaborations with both academic and industrial R&D partners. Beyond research, MLIA contributes significantly to community service and plays an active role in the administrative aspects of research and teaching within Sorbonne University. Their expertise spans areas like computational intelligence, artificial learning, explainable AI (XAI), and hybrid AI, making them a key contributor to advancements in the field.

CiteBuddy is an AI-powered tool designed to streamline legal citation management directly within Microsoft Word. It offers features like intelligent footnote suggestions, citation verification across multiple styles (OSCOLA, Bluebook, Harvard, APA, Chicago), and smart PDF analysis. Users can upload and analyze PDFs to extract key information and generate accurate citation suggestions. The tool aims to help law students and professionals master legal citations, ensuring flawless documents. It provides various pricing tiers, including options for bulk footnote formatting, citation history tracking, and team collaboration, making it suitable for individual users and research teams alike.