Fujitsu and MoBagel revolutionize business processes through accelerated AI prediction

How the Integrated AI Solution Works

The Fujitsu-MoBagel collaboration embeds three interconnected features within Decanter AI. First, the Data Verification and Anomaly Scoring system evaluates input data quality and relevance, categorising datasets as 'Healthy', 'Moderate', or 'Critical' to guide refinements before modelling. This automated screening reduces flawed predictions from poor data quality, a common AI implementation barrier. Second, the Algorithm Recommendation System uses Fujitsu Kozuchi AutoML to analyse data characteristics and suggest optimal machine learning models for specific prediction tasks, eliminating guesswork and accelerating model selection. Third, Feature Analysis powered by Wide Learning™ technology identifies influential variables post-modelling using weighted combinations to illustrate feature importance. These insights are interpreted by a large language model and presented in plain language, enabling non-technical business users to understand and act on AI-generated findings.

A practical example: predicting car insurance purchase likelihood. The system scans historical data for anomalies like duplicate entries or implausible values, then recommends an optimal model—such as random forest—for the task. It analyses which customer attributes (age, location, vehicle type) most strongly correlate with purchase intent, assigns weighted importance, and delivers a clear summary explaining why certain profiles are more likely to buy insurance. This end-to-end process, from raw data to actionable insight, occurs significantly faster than traditional AI development cycles while maintaining statistical rigour.

The integrated approach enables model recommendations to be generated up to four times faster than previous methods without compromising accuracy. This speed improvement comes from automating iterative tasks that typically consume substantial data science time, such as algorithm testing and feature engineering. The solution also emphasises explainability, addressing a key concern in enterprise AI adoption where stakeholders require transparency into prediction generation before trusting or acting on results.

Business Impact and Real-World Applications

The partnership targets practical business challenges across finance, insurance, retail, and manufacturing. MoBagel notes its platform has assisted over 11,000 brands in deploying AI for sales forecasting, supply chain optimisation, financial risk assessment, and production efficiency improvements. By reducing AI implementation timelines from 12 months to approximately one month, businesses can respond more swiftly to market changes, test hypotheses rapidly, and deploy iterative improvements without prolonged development cycles.

A forthcoming proof of concept with Deloitte Touche Tohmatsu Limited will explore the technology's application in financial crime prevention. In this workshop-based initiative, participants will use Decanter AI to analyse large volumes of trading data to identify anomalous patterns indicative of insider trading, money laundering, or fraudulent activities. Rapid processing and interpretation of complex financial datasets could significantly enhance monitoring capabilities for banks, regulators, and financial institutions seeking to detect sophisticated illicit behaviours.

Beyond specific use cases, the collaboration reflects a broader trend toward democratising AI access. By packaging sophisticated machine learning capabilities into a guided, automated workflow, Fujitsu and MoBagel aim to lower technical barriers that have historically limited AI adoption to large corporations with dedicated data science teams. Small and medium-sized enterprises, including those in Nigeria's growing digital economy, could leverage such tools to gain predictive insights previously accessible only to well-resourced organisations. The emphasis on cost-effectiveness and rapid deployment aligns with the needs of businesses in competitive, fast-moving environments where delayed insights can translate to missed opportunities.

Strategic Context and Future Outlook

The partnership builds on a strategic alliance formed between Fujitsu and MoBagel in December 2023, combining established enterprise technology capabilities with agile startup innovation. Fujitsu positions its Kozuchi service as a vehicle for delivering specialised AI functionalities—such as AutoML and explainable AI—to external partners, extending the reach of its research and development investments. For MoBagel, access to Fujitsu's proven technologies enhances platform credibility and capabilities, particularly in regulated industries where explainability and auditability are essential.

Hideto (Ted) Okada, Senior Vice President and Head of Technology Strategy Unit at Fujitsu Limited, characterised the collaboration as accelerating customer AI adoption while fostering innovation in the broader AI ecosystem. By supplying technology to startups like MoBagel, Fujitsu aims to stimulate novel applications that might not emerge from internal development alone. Adms Chung, CEO of MoBagel Inc., highlighted the synergy between partners, noting the integration addresses widespread industry pain points related to the trade-off between model accuracy, development speed, and implementation cost.

Looking ahead, both companies plan to expand the solution's availability and applicability. Fujitsu intends to continue offering its Kozuchi AI services to partners seeking to build differentiated AI offerings, while MoBagel plans to pursue additional industry-specific proofs of concept and expand its global footprint through offices in Tokyo, Taipei, Singapore, and other locations. The long-term vision centres on creating reusable AI components that can be rapidly assembled into bespoke solutions for distinct business problems, moving away from custom-built models toward more standardised, configurable AI services.

What this means

The Fujitsu-MoBagel partnership represents a pragmatic step toward making AI prediction capabilities more accessible, transparent, and timely for business users. By combining Fujitsu's enterprise-grade AI technologies with MoBagel's automated platform approach, the collaboration addresses persistent obstacles in AI implementation: lengthy development cycles, technical complexity, and unclear model outputs. The resulting solution offers a structured pathway from raw data to interpretable predictions, with built-in safeguards for data quality and model suitability. For organisations in Nigeria and globally grappling with how to harness AI without excessive investment in specialised talent or infrastructure, such integrated tools provide a viable starting point. As the proof of concept with Deloitte demonstrates, the technology's value extends beyond theoretical applications to concrete use cases in risk monitoring and decision support. While no single solution can eliminate all challenges associated with AI adoption, innovations that streamline core workflows while maintaining explainability and accuracy contribute meaningfully to the broader goal of enabling data-driven transformation across industries.