Research & Education

KAIST AI is the Kim Jaechul Graduate School of AI at the Korea Advanced Institute of Science and Technology, providing comprehensive academic programs in Artificial Intelligence. The institution offers various degree options, including Master of Science (MS), integrated MS and PhD, and Doctor of Philosophy (PhD) degrees, focusing on cutting-edge AI research. Its distinguished faculty members are experts in diverse fields such as Machine Learning, AI, Data Mining, Computer Vision, Natural Language Processing (NLP), and Robotics. KAIST AI is dedicated to fostering advanced research and education, preparing students for leadership roles in the rapidly evolving AI landscape.

Griphic is an AI development platform founded by IIT Delhi engineers, dedicated to enriching lives through technological innovation. The company focuses on harnessing advanced technologies to address real-world challenges and drive industry disruption. Its core technologies include Extended Reality (XR), Machine Learning (ML), Artificial Intelligence (AI), Interactive Interfaces, 3D Visualization, Large Language Models (LLMs), Computer Vision, and Rigging and Animation. Griphic's team comprises talented engineers, AI/ML specialists, and VR developers, ensuring comprehensive and innovative solutions. The company is backed by directors of SKETS Studio, a leading design documentation firm specializing in BIM, architecture, VR, and 3D visualization, providing strong industry support and expertise.

Qubitron AI is a forward-thinking artificial intelligence consultancy dedicated to transforming businesses through intelligent automation and data-driven innovation. They offer a range of services including AI strategy and architecture, workflow automation systems, custom LLM and AI systems, data engineering and analytics, and integration, scaling, and deployment. Their mission is to empower organizations with scalable, effective, and modern artificial intelligence solutions, helping them innovate faster and operate smarter. Qubitron AI focuses on delivering production-ready AI systems that scale operations and accelerate business performance, with a strong emphasis on client satisfaction and tailored solutions.

Catalytic Data Science offers an AI-powered platform designed for regulated workflows within the life sciences industry. The platform, Catalytic Intelligence, integrates AI, data, and workflows to deliver actionable insights and support informed decision-making. It features a leading content ingestion engine that transforms regulatory precedent information into a comprehensive multinational data foundry, which can also be applied to internal regulatory content. Key capabilities include an intuitive conversational interface for complex AI workflows, multi-agent workflows for powerful analysis, and a library of specialized AI agents validated on life science data. The platform aims to make regulatory strategy predictive, automate report creation, and enable faster data analysis for deeper insights.

Nano Banana Pro is an AI image generation and editing tool designed for creators and marketers. It allows users to generate high-quality images from text prompts and offers advanced controls for photo editing. The platform emphasizes fast results, character consistency, and seamless image blending, making it suitable for various creative needs. Users can start with free credits, with options for higher resolutions and commercial use available on paid plans. The tool supports JPEG/PNG uploads up to 10MB and delivers high-quality JPEG downloads. API access is also available for higher-tier plans, catering to more integrated workflows.

PaiByTwo Pvt. Ltd. is a deep tech robotics startup founded by research scholars at IIT Roorkee, specializing in Artificial Intelligence and Robotics. The company focuses on providing opportunities to enhance AI learning through research and development, offering quality training in key AI pillars such as Robotics, Computer Vision, Emotion AI, Machine Learning, Deep Learning, and Statistics. They cater to a wide range of learners, from school students to working professionals, through their 'PaiByTwo Classroom' initiative. Additionally, 'PaiByTwo Labs' innovates in cyber-physical systems, aiming to develop emotionally intelligent robots and provide services for agriculture, healthcare, smart cities, and smart manufacturing, aligning with Industry 4.0 readiness.

Zelthy is an AI-native application platform designed for life sciences operations, founded in 2017. It empowers pharmaceutical companies to rapidly deploy patient support programs, advanced therapy orchestration, commercial operations, traceability, compliance, and regulatory workflows within 4-8 weeks. Built on Zango, an open-source Django framework for regulated industries, Zelthy offers full code transparency and can be deployed on cloud, on-premise, or hybrid environments. The platform includes AI agents for custom workflow automation, pre-built templates for over 45 life sciences use cases, and robust features like RBAC, audit logging, and multi-tenancy. Trusted by leading pharma companies, Zelthy supports over 300 production applications across 12+ countries, impacting over 1 million patients.

Machina Sports is a comprehensive platform designed to accelerate the development and deployment of AI solutions in the sports industry. It acts as an operating system for sports AI, providing live data, real-time intelligence, and production-ready agents. The platform simplifies the process of teaching AI about sports leagues, rosters, and fixtures, managing real-time data pipelines, and integrating with various data sources through a single API call. It offers features like live sports data streaming, sports-native AI agents, real-time analytics, fan personalization tools, and multimodal output generation for highlights and podcasts. Machina Sports aims to eliminate the complexities of in-house AI infrastructure, allowing developers and rights holders to ship features rapidly.

MOLCURE Inc. is an AI company dedicated to accelerating drug discovery by integrating machine learning, biotechnology, and robotics. Their core technology is an evolving AI platform featuring specialized algorithms for biological data and a rapid in-house learning loop. Key strengths include a proprietary Antibody Large Language Model (LLM) built from scratch, a database with over 1 billion proprietary datapoints, and an in-house wet lab for directed evolution experiments and next-generation sequencing. MOLCURE offers various collaboration models, including end-to-end support for AI-driven molecular design, AI-powered de novo discovery of novel molecules, and AI-driven lead optimization for existing candidate molecules. They also provide consulting services to leverage AI technologies in drug discovery processes.

Bluenote is an advanced AI platform designed for the life sciences industry, specifically to accelerate regulatory submissions by 50-75%. Trusted by leading pharmaceutical, medical device, diagnostics, CROs, and CDMOs, Bluenote's Agentic AI platform automates critical documentation with high accuracy and enterprise-grade security. Key features include instant generation of scientific and regulatory documents aligned with templates and global guidelines, an AI Assistant for refining data presentations and running gap analyses, and AI Agents that automate repetitive, multi-step workflows. The platform also offers robust dataset search capabilities and technical translation services, all grounded in user data and standards with 100% traceability. It supports enterprise-wide scalability and integrates with existing data sources like ELN and QMS/DMS.

CVisionLab is a leading R&D service provider specializing in custom computer vision and AI solutions. They offer outsourcing services for developing bespoke computer vision applications powered by machine learning and artificial intelligence. Their expertise spans various domains, including face analysis, NSFW recognition, brand recognition, object detection, image labeling, mask detection, and PPE recognition. CVisionLab also provides cloud solutions as an official Google Cloud and Yandex Cloud Partner, enabling end-to-end AI/ML solution deployment. With over 10 years of experience and a focus on cutting-edge technologies, they serve a diverse clientele from healthcare to agriculture, offering transparent workflows and professional development teams.

Deargen is an AI-first pharmatech company dedicated to transforming the future of treating undruggable diseases through the integration of AI and collective intelligence. Their Dr. UG AI platform leverages deep learning technology for genome data analysis, biomarker prediction, molecule selection, and optimization. Deargen discovers new disease targets and designs small molecules for novel drugs, enabling precision medicine. Their AI technology is applied across various new drug development strategies, including PROTAC, ADC, EPITOPE, AGONIST, and INHIBITOR pipelines.

Sketch is an AI code-writing assistant tailored for pandas users, designed to understand the context of your data to provide highly relevant code suggestions. It offers a natural language interface, allowing users to ask questions and generate code blocks for various data analysis tasks. Sketch supports data cataloging, cleaning, feature creation, analysis, and visualization. It can be integrated into a workflow by simply importing the library and using the `.sketch` extension on any pandas DataFrame. Users can leverage its `ask`, `howto`, and `apply` functions for querying data, generating code, and parsing fields or creating new features, respectively. Sketch uses efficient approximation algorithms to summarize data for language models and can run locally with pre-built Hugging Face models or connect to OpenAI directly.

Nurjana Technologies & Space is an established industry player offering highly specialized engineering solutions globally, particularly for defense, space, and security sectors. They are pioneers in developing precise, reliable solutions that integrate the latest sensor technology, artificial intelligence, and machine learning with state-of-the-art hardware. Their capabilities include command and control systems, optical tracking, electro-optical systems, AI video tracking, and smart drones. They also focus on electromagnetic field management, critical infrastructure protection, border control, and wildfire detection. NurjanaTech leverages AI for real-time identification, tracking, recording, and information sharing to enhance mission effectiveness, providing advanced software solutions for situational awareness and informed decision-making in complex environments.

RePAIRproject2021, an acronym for Reconstructing the Past: Artificial Intelligence and Robotics meet Cultural Heritage, is a research initiative focused on revolutionizing archaeological reconstruction. The project employs state-of-the-art AI and robotics technology to physically reassemble fragmented archaeological artifacts, such as vases, amphorae, and frescoes. Its primary goal is to develop innovative technology that significantly reduces the time and effort traditionally required for reconstructing shattered works of art, which are often found in numerous pieces at excavation sites. This project addresses the challenge of reassembling damaged, worn out, or entirely missing fragments, aiming to make the process more efficient and less frustrating for researchers.

AdaLab.ai is an AI research and product studio based in Hamburg, specializing in building and shipping machine learning products. They offer custom ML solutions for enterprise clients, tackling complex problems and putting the latest research into production. A notable product from their studio is Plazmapunk, an AI-powered music visualizer that transforms any track into a music video without requiring editing skills. Plazmapunk features AI-powered visuals that move to the beat, an intuitive interface, and real-time preview capabilities, allowing users to tweak and iterate quickly. AdaLab.ai emphasizes a practical approach, focusing on delivering working AI products rather than just concepts.

ctransformers is a Python library designed to provide efficient bindings for Transformer models, leveraging C/C++ implementations and the GGML library. This allows for optimized local execution of various large language models, including GPT-2, LLaMA, Falcon, and more. The library offers a unified interface for model loading and generation, supporting both local files and Hugging Face Hub repositories. It integrates seamlessly with popular AI frameworks like Hugging Face Transformers and LangChain, making it versatile for different development workflows. ctransformers also provides GPU acceleration for compatible models, with support for CUDA, Metal, and ROCm, and includes experimental GPTQ quantization support for LLaMA models, enhancing performance and reducing memory footprint.

Agrematch is an artificial intelligence company specializing in the discovery and development of sustainable bioactive compounds for the agriculture and food sectors. Its disruptive AI predictive system, AI4AI, leverages machine/deep learning algorithms and a proprietary database of billions of compounds to identify molecules with biological activity. The platform predicts and characterizes crucial compound attributes like toxicity, mode of action, compatibility, and environmental impact early in the screening process, accelerating development and mitigating downstream risks. Agrematch's unique approach allows for the rapid identification and validation of novel products, offering a significant advantage over traditional discovery methods.

DHARMA-AI is an AI Lab specializing in the development of Specialized Small Language Models (SLMs) using proprietary hard tech. Their mission is to transform AI by creating intelligent, personalized, and efficient solutions for businesses. They focus on delivering custom AI models trained with client-specific data, leading to significant cost reductions (up to 50 times less than large market models) and improved accuracy. DHARMA-AI's approach prioritizes ethical, sustainable, and secure AI, with SLMs consuming 30 to 50 times less energy than traditional LLMs. They offer various business lines, including ready-to-use AI products like Smart OCR and KM Maintenance, and an AI Lab for custom solution development and MLOps support.

ORRECO is an advanced AI platform designed for professional sports, leveraging cutting-edge AI alongside human guidance to accelerate recovery, optimize performance, and prolong athletes' careers. The platform unifies staff workflows, data, and AI agents, offering customizable modules that scale seamlessly with a purpose-built athlete app. Key features include objective blood biomarkers for real-time decision-making, centralized medical notes, holistic recovery labs, and computer-vision movement trends to flag non-contact soft-tissue injury risk early. ORRECO also provides specialized modules for women's performance, incorporating menstrual cycle tracking and expert advice. With over 20 years of experience supporting Olympic athletes and elite teams, ORRECO bridges the gap between lab precision and on-field performance, delivering measurable results like significant reductions in injuries and lost training days.

Captum is an open-source PyTorch library designed for model interpretability, allowing users to understand the predictions of their AI models. It supports a wide range of modalities, including vision and text, making it versatile for different types of machine learning applications. Built specifically for PyTorch, Captum integrates seamlessly with most PyTorch models, requiring minimal modifications to existing neural networks. The library is extensible, providing a generic framework for interpretability research, which enables developers and researchers to easily implement and benchmark new algorithms. Captum is a valuable tool for anyone looking to gain deeper insights into their PyTorch models' decision-making processes.

Terray Therapeutics is at the forefront of small molecule drug discovery, utilizing a unique integrated full-stack AI platform. The company's approach emphasizes the synergy between wet lab science and advanced AI, generating high-quality data at an unprecedented scale. Their experimental dataset boasts over two billion unique target-ligand binding measurements, expanding by one billion quarterly. This precise, data-rich environment enables accurate molecular property prediction through deep learning regression models. Terray's closed-loop iteration process allows for a design-make-test-analyze cycle in less than a month per target, efficiently exploring chemical space and identifying potent, selective small molecules for various targets in parallel. This platform is designed to significantly improve the speed, cost, and success rate of drug discovery and development, with an internal pipeline focused on immunology and partnerships across diverse therapeutic areas.

WEO creates environmental analytics from satellites, leveraging the power of space data and AI to generate actionable insights. The platform focuses on making cities more resilient, streamlining vegetation management, and reducing environmental risks. WEO utilizes cutting-edge machine learning techniques to enhance the resolution of satellite imagery, providing detailed data for clients. Key applications include monitoring impermeable surfaces for water cycle restoration, supporting climate change risk analysis for cities, and deploying satellite- and AI-based technology to quantify heatwave impacts for parametric insurance. WEO's mission is to work towards a sustainable future with impactful Earth Observation environmental solutions for all.

Vulcan Augmetics develops AI-powered myoelectric prosthetic hands and wireless EMG sensors, designed to provide a complete upper-limb prosthetic solution. The system uses machine learning to enable fast adaptation and precise control for users with upper limb loss, with some patients adapting in under 30 minutes. It offers smart and fast signal checks, clear-cut fabrication, and easy, accurate fitting with one-minute auto-calibration. The Vulcan Hand supports over 30 practical daily activities, and the Vulcan Control Solution effectively detects signals even on complex limb cases. The technology is patient-proven and aims to provide high-fidelity biodata and feedback for healthcare professionals and patients, with continuous clinical support and a presence in over 17 countries.