Bridging Understanding and Implementation
cognisspaer was founded on the recognition that many organisations understand the potential value of machine learning but face practical barriers to implementation. Our mission is to provide clarity and capability in equal measure.
Return Home
Our Story
cognisspaer was established in 2019 by a group of machine learning practitioners and domain specialists who had observed a recurring pattern: organisations with legitimate use cases and adequate resources still struggled to move from concept to deployment. The barriers were rarely technical in nature. More often, they involved gaps in shared understanding, unclear success criteria, or misalignment between technical possibilities and operational realities.
Our team brought experience from financial services, academic research, and technology consulting. What united us was frustration with implementations that failed not because the technology was inadequate, but because the groundwork had not been properly laid. We recognised that many organisations needed a partner who could speak both the language of machine learning and the language of business constraints.
From the beginning, we committed to several principles. First, we would not oversell capabilities or promise outcomes we could not reasonably deliver. Second, we would invest time upfront in understanding the actual problem rather than rushing to apply predetermined solutions. Third, we would build systems with explainability and auditability as core requirements, recognising that many of our clients operate in regulated environments.
Today, cognisspaer operates primarily in Singapore, working with financial institutions, professional services firms, and innovation teams. Our engagements range from rapid prototyping sprints that validate technical feasibility to comprehensive implementations that integrate with existing operational systems. We remain focused on the intersection of technical capability and practical deployment.
Our Team
Specialists combining technical depth with practical implementation experience
Dr. David Koh
Principal Consultant
Brings fifteen years of experience in machine learning applications for financial services. Previously led data science teams at major regional banks, focusing on credit risk modelling and regulatory compliance frameworks.
Mei Ling Tan
Technical Lead
Specialises in natural language processing and semantic knowledge systems. Holds advanced degrees in computational linguistics and has developed extraction pipelines for legal and research organisations across Southeast Asia.
Arun Ramesh
Implementation Specialist
Focuses on system integration and deployment architecture. Background in enterprise software development with particular expertise in ensuring AI systems meet production readiness and operational sustainability requirements.
Professional Standards
Our approach to quality assurance and compliance in AI development
Model Documentation
Every model we develop includes comprehensive technical documentation covering architecture decisions, training methodology, validation procedures, and performance characteristics. Documentation is structured to support both technical review and regulatory audit requirements.
Data Protection
We adhere to Singapore's Personal Data Protection Act and implement appropriate technical and organisational measures for data handling. This includes encryption protocols, access controls, data minimisation principles, and clear retention policies.
Validation Framework
All implementations undergo structured validation against defined success criteria. This includes performance testing on held-out data, bias assessment, edge case analysis, and verification that model behaviour aligns with intended use cases and constraints.
Explainability
We prioritise model interpretability, particularly for applications in regulated sectors. Implementations include mechanisms for understanding individual predictions, feature importance analysis, and documentation of decision logic suitable for stakeholder review.
Version Control
Rigorous version management for code, data, and models ensures reproducibility and supports rollback capabilities. We maintain clear lineage documentation tracking changes from development through deployment, essential for audit trails.
Ethical Guidelines
Our development process incorporates fairness assessments, privacy impact analysis, and consideration of broader societal implications. We engage with clients on responsible AI principles and help structure governance frameworks appropriate to their context.
Values and Approach
Our work is grounded in the understanding that machine learning systems exist within larger organisational and societal contexts. Technical excellence matters, but so does alignment with operational realities, regulatory requirements, and human decision-making processes. We invest significant effort in the foundational work that precedes implementation: problem formulation, data assessment, stakeholder alignment, and success criteria definition.
We recognise that AI systems are tools rather than solutions in themselves. The value comes from how they are integrated into workflows, how their outputs inform decisions, and how they complement rather than replace human judgment. Our engagements focus on this integration challenge as much as on the technical development of models.
Transparency is central to our practice. We maintain open communication about technical limitations, provide realistic timelines, document our decision-making processes, and ensure clients have clear understanding of what has been delivered and how to maintain it. This approach builds trust and enables organisations to make informed decisions about their AI investments.
Ready to Explore What's Possible?
We welcome conversations with organisations considering machine learning implementations. Reach us to arrange an initial discussion about your requirements.
Contact Our Team