The Client |
|
This client is a Singapore-headquartered cybersecurity leader, renowned for its deep-tech solutions in software composition analysis, vulnerability detection, and AI application security. With origins in cutting-edge research from Nanyang Technological University (NTU), the company has collaborated with major clients such as Singtel, Apple, Adobe, and the National University of Singapore. It has been recognized with multiple awards including the TechBlazer Award for Most Promising Innovation (2023).
The company is one of the few globally working on AI security for Large Language Models (LLMs)—a critical and emerging field given the rapid adoption of AI in commercial applications. Their proprietary solutions, ModelScan and ModelDefender, target vulnerabilities like prompt injection, model poisoning, and anomaly detection in generative AI models. They currently serve clients across Southeast Asia and are actively expanding into Japan through a high-impact partnership with a leading Japanese AI company.
The company is one of the few globally working on AI security for Large Language Models (LLMs)—a critical and emerging field given the rapid adoption of AI in commercial applications. Their proprietary solutions, ModelScan and ModelDefender, target vulnerabilities like prompt injection, model poisoning, and anomaly detection in generative AI models. They currently serve clients across Southeast Asia and are actively expanding into Japan through a high-impact partnership with a leading Japanese AI company.
How RIC Helped |
|
Grant Secured:
RIC supported the client in securing the Enterprise Development Grant (EDG) – Co-Innovation Programme (CIP) under the Eureka Globalstars-Japan Call, receiving close to S$200,000 in funding to support up to 70% of eligible development costs.
Scope of RIC’s Involvement:
RIC provided comprehensive project scoping and advisory support, including:
The solutions involved cutting-edge cybersecurity measures tailored to enterprise AI Large Language Model (LLM) systems—a space still lacking standardized frameworks. The R&D activities proposed were unprecedented in complexity, combining static code analysis, malicious prompt filtering, and real-time anomaly detection
RIC supported the client in securing the Enterprise Development Grant (EDG) – Co-Innovation Programme (CIP) under the Eureka Globalstars-Japan Call, receiving close to S$200,000 in funding to support up to 70% of eligible development costs.
Scope of RIC’s Involvement:
RIC provided comprehensive project scoping and advisory support, including:
- Strategic grant matching and proposal development, identifying the EDG (CIP) under the Japan-Singapore innovation corridor as the optimal funding route.
- Deep technical articulation of the cybersecurity solutions being developed (ModelScan and ModelDefender) targeting vulnerabilities in Japanese LLM applications.
- Detailed documentation, including project milestones, cost tables, and projected quantitative and qualitative outcomes.
- Market analysis and risk management, assessing Japan’s US$7.25 billion cybersecurity market and justifying expansion into a technically sophisticated, regulation-sensitive environment.
- Collaboration planning with a Japanese AI firm, detailing integration strategies and co-development structure to align with cross-border innovation requirements.
- In addition to our strategic and technical consulting, RIC also played a critical role in navigating the complex Eureka Globalstars portal, which was required for this cross-border grant application. Unlike Singapore’s Business Grants Portal (BGP), the Eureka platform is an international system with extensive technical sections, detailed co-signature requirements, and rigid formatting constraints, making it notoriously difficult to manage without experience. RIC guided the applicant on all aspect—from coordinating cross-partner documentation, crafting and uploading the Eureka submission, to ensuring compliance with both Singaporean and Japanese requirements. This ensured the proposal not only met the standards of Enterprise Singapore but also aligned seamlessly with Eureka’s international evaluation framework. Our ability to manage such multi-jurisdictional grant systems demonstrates RIC’s readiness to support clients in complex global innovation funding programs.
The solutions involved cutting-edge cybersecurity measures tailored to enterprise AI Large Language Model (LLM) systems—a space still lacking standardized frameworks. The R&D activities proposed were unprecedented in complexity, combining static code analysis, malicious prompt filtering, and real-time anomaly detection
Outcomes |
|
The project, currently in execution, is expected to deliver:
RIC worked hand-in-hand with the client throughout the process—from concept to grant approval to execution support. Our technical understanding of AI and cybersecurity, combined with fluency in innovation grants, was key in framing this complex project in a way that secured government support and set the client on a clear path to commercialization and expansion.
This case underscores RIC’s strengths in delivering outcomes for deep-tech innovators in emerging fields. With an unique ability to scope, articulate, and support high-risk, high-reward projects, RIC is the consultancy of choice for companies at the frontier of AI, cybersecurity, and innovation.
- Two novel AI cybersecurity products: ModelScan (static analysis tool for AI apps and models) and ModelDefender (prompt filtering and anomaly detection engine).
- Market-ready tools for deployment in enterprise LLM applications, with significant implications for banks, governments, and healthcare providers developing their own generative AI models.
- Over S$450,000 in annual productivity savings for typical enterprise clients, due to reduced DevSecOps man-hours across vulnerability scanning, penetration testing, and access control.
- Strategic market entry into Japan, positioning the company to gain up to 2.6% market share in the LLM cybersecurity segment within five years.
- Enhanced internal capabilities, including upskilling in regulatory compliance, anomaly detection, and secure software lifecycle processes.
- Global thought leadership, as demonstrated by multiple research publications authored by the company’s team on adversarial AI and LLM jailbreak vulnerabilities.
RIC worked hand-in-hand with the client throughout the process—from concept to grant approval to execution support. Our technical understanding of AI and cybersecurity, combined with fluency in innovation grants, was key in framing this complex project in a way that secured government support and set the client on a clear path to commercialization and expansion.
This case underscores RIC’s strengths in delivering outcomes for deep-tech innovators in emerging fields. With an unique ability to scope, articulate, and support high-risk, high-reward projects, RIC is the consultancy of choice for companies at the frontier of AI, cybersecurity, and innovation.