Aazos Launches Support for SPReAD Applicants Using 'Multi-Sigma®' AI Platform – Optimizing Experiments with Small Data and No-Code, Including Patented Causal Analysis
Aazos Inc. has started offering specialized support for researchers applying for the MEXT 'SPReAD' program. The support leverages their 'Multi-Sigma®' no-code AI platform to assist from proposal writing to post-adoption research optimization over a 6-month period.
📋 Article Processing Timeline
- 📰 Published: April 28, 2026 at 23:26
- 🔍 Collected: April 28, 2026 at 15:02
- 🤖 AI Analyzed: April 28, 2026 at 15:10 (8 min after Collected)
■ Key Points of this Release
AI Utilization Support for SPReAD Applicants
Aazos provides consistent support for researchers, from application preparation to AI environment construction after acceptance.
Support for Research Timeline and Application Requirements
Aazos provides step-by-step AI utilization support aligned with the approximately 6-month research plan, specifically addressing the application requirement for 'extraction and sharing of AI know-how.'
Aazos Inc. (Tsukuba, Ibaraki; CEO: Kotaro Kawajiri) has launched individual support services utilizing 'Multi-Sigma®,' a no-code R&D AI analysis platform capable of handling small datasets, for researchers considering applying for the Ministry of Education, Culture, Sports, Science and Technology's (MEXT) 'AI for Science Innovation Program: Exploratory Challenge Research Creation Project (SPReAD)' (1st application deadline: May 18, 2026; 2nd application expected in early June). The service provides end-to-end support, from building the AI analysis environment required after selection to addressing the unique application item: 'Plan for Realizing AI Know-how Extraction and Sharing.'
■ Background: Computing Resources Alone Do Not Drive Research
SPReAD is a research support system with a scale of up to 5 million yen per project and approximately 1,000 projects annually. The research period is about 6 months, from selection until January 6, 2027. It targets experimental researchers in all fields and is open to those who are using AI seriously for the first time in their research.
Securing cloud computing resources is an important first step, but to actually advance research, direct intervention in the process is required, such as 'selecting AI methods suited to the theme,' 'predicting from small data,' and 'proposing experimental conditions to try next.' Aazos provides Multi-Sigma® utilization support as a means to bridge this gap.
■ Multi-Sigma®: An AI Analysis Platform Developed by Experimental Researchers for Experimental Researchers
Multi-Sigma® is an R&D AI analysis platform that provides no-code services for prediction, Bayesian optimization, multi-objective optimization, causal factor analysis, causal chain analysis, and experimental design on the cloud. Founder Kotaro Kawajiri developed the prototype based on challenges he faced during his 18 years of experimental research at AIST, specifically in nanoweap-particle manufacturing using plasma equipment (7 parameters, 10 million condition spaces, and multi-objective trade-offs).
Key Features:
Small Data Compatibility: Enables prediction and optimization from limited experimental data using Gaussian process regression and neural networks.
No Coding Required: Researchers can start operations themselves from the day of selection.
Bayesian Optimization: AI suggests the 'next experimental conditions to try,' making exploration efficient.
Multi-objective Optimization: Simultaneously solves competing design requirements (e.g., strength vs. weight, yield vs. purity).
Causal Chain Analysis (International Patent Pending): The world's first no-code commercial platform implementation for analyzing causal structures across multiple experimental processes.
AI Sharing Function: Share trained models on the platform so other users can utilize them directly.
Experimental Design: Extracts maximum information from a limited number of experiments.
Since its release in 2021, Multi-Sigma® has established a solid position with over 380 corporate and university users. It is used across various fields such as materials, drug discovery, machinery, agriculture, and social sciences, promoting interdisciplinary and inter-organizational collaboration.
■ How to Run a 6-Month Research Plan with Multi-Sigma®
Phase 1 (Months 1-2): Research Design & Initial Data Collection
Create experimental designs based on research goals. Acquire experimental data.
Phase 2 (Month 3): Model Building, Verification, & Optimization
Build prediction models with available data, verify accuracy, and explore optimal solutions through factor analysis.
Phase 3 (Months 4-5): Verification of Optimal Conditions & Additional Data
Verify optimal solutions and perform re-analysis with additional data.
Phase 4 (Month 6): Result Summary & Know-how Organization
Submit papers and organize insights for sharing.
AI Utilization Support for SPReAD Applicants
Aazos provides consistent support for researchers, from application preparation to AI environment construction after acceptance.
Support for Research Timeline and Application Requirements
Aazos provides step-by-step AI utilization support aligned with the approximately 6-month research plan, specifically addressing the application requirement for 'extraction and sharing of AI know-how.'
Aazos Inc. (Tsukuba, Ibaraki; CEO: Kotaro Kawajiri) has launched individual support services utilizing 'Multi-Sigma®,' a no-code R&D AI analysis platform capable of handling small datasets, for researchers considering applying for the Ministry of Education, Culture, Sports, Science and Technology's (MEXT) 'AI for Science Innovation Program: Exploratory Challenge Research Creation Project (SPReAD)' (1st application deadline: May 18, 2026; 2nd application expected in early June). The service provides end-to-end support, from building the AI analysis environment required after selection to addressing the unique application item: 'Plan for Realizing AI Know-how Extraction and Sharing.'
■ Background: Computing Resources Alone Do Not Drive Research
SPReAD is a research support system with a scale of up to 5 million yen per project and approximately 1,000 projects annually. The research period is about 6 months, from selection until January 6, 2027. It targets experimental researchers in all fields and is open to those who are using AI seriously for the first time in their research.
Securing cloud computing resources is an important first step, but to actually advance research, direct intervention in the process is required, such as 'selecting AI methods suited to the theme,' 'predicting from small data,' and 'proposing experimental conditions to try next.' Aazos provides Multi-Sigma® utilization support as a means to bridge this gap.
■ Multi-Sigma®: An AI Analysis Platform Developed by Experimental Researchers for Experimental Researchers
Multi-Sigma® is an R&D AI analysis platform that provides no-code services for prediction, Bayesian optimization, multi-objective optimization, causal factor analysis, causal chain analysis, and experimental design on the cloud. Founder Kotaro Kawajiri developed the prototype based on challenges he faced during his 18 years of experimental research at AIST, specifically in nanoweap-particle manufacturing using plasma equipment (7 parameters, 10 million condition spaces, and multi-objective trade-offs).
Key Features:
Small Data Compatibility: Enables prediction and optimization from limited experimental data using Gaussian process regression and neural networks.
No Coding Required: Researchers can start operations themselves from the day of selection.
Bayesian Optimization: AI suggests the 'next experimental conditions to try,' making exploration efficient.
Multi-objective Optimization: Simultaneously solves competing design requirements (e.g., strength vs. weight, yield vs. purity).
Causal Chain Analysis (International Patent Pending): The world's first no-code commercial platform implementation for analyzing causal structures across multiple experimental processes.
AI Sharing Function: Share trained models on the platform so other users can utilize them directly.
Experimental Design: Extracts maximum information from a limited number of experiments.
Since its release in 2021, Multi-Sigma® has established a solid position with over 380 corporate and university users. It is used across various fields such as materials, drug discovery, machinery, agriculture, and social sciences, promoting interdisciplinary and inter-organizational collaboration.
■ How to Run a 6-Month Research Plan with Multi-Sigma®
Phase 1 (Months 1-2): Research Design & Initial Data Collection
Create experimental designs based on research goals. Acquire experimental data.
Phase 2 (Month 3): Model Building, Verification, & Optimization
Build prediction models with available data, verify accuracy, and explore optimal solutions through factor analysis.
Phase 3 (Months 4-5): Verification of Optimal Conditions & Additional Data
Verify optimal solutions and perform re-analysis with additional data.
Phase 4 (Month 6): Result Summary & Know-how Organization
Submit papers and organize insights for sharing.