Overview

100 Experiments (100E) is AI Singapore's flagship initiative to drive AI adoption in industry by co-creating AI solutions with companies. The goal is to allow organisations to transform real-world business problems into AI solutions (proofs of concept or production-ready models) while building internal capability. AI Singapore provides in-house AI engineering teams, AI Apprentices (from the AI Apprenticeship Programme, AIAP®), and access to infrastructure, as part of collaborative projects.

Support Details

Support Rate

AI Singapore provides co-funding support (grants part of the cost) with organisation matching via in-kind and cash contributions

Cap

Up to S$150,000 per project with matching contributions required from organisation

Project Duration & Scope

6 months for MVP path or 3 months for PoC variant, structured in phases: assessment/scoping → development → handover/knowledge transfer

Eligibility Criteria

Organisations must propose a real business challenge (not a generic problem) where no off-the-shelf AI solution fully meets the need
Data readiness is evaluated (quality, volume, access) - projects with poor datasets may be rejected or downgraded
Must provide in-kind contributions (staff time, domain experts) and potentially cash contributions as part of project cost matching
Projects undergo multi-stage assessment including technical feasibility, business impact, AI governance compliance
Strong emphasis on AI governance/responsible AI, data ethics, model explainability, oversight
100E adheres to the Model AI Governance Framework
Alignment with corporate/national priorities required

Supported Categories

100E (3-month POC): Shortened, intensive path to validate an AI concept quickly with same methodology and team support

100E (6-month MVP/standard): More time to build, refine, and deliver a production-ready AI solution with greater depth and integration

Application Process

Scoping and assessment - consult with AI Singapore to assess AI readiness via AIRI or internal scoping

Define business use case, problem statement, data readiness

Develop baseline/initial model or prototype to test viability

Submit technical proposal with resource plan, milestones, team roles, KPIs

Undergo AI Singapore's evaluation (engineering, governance, business impact)

Implementation and development with agile sprint cycles

Iterate models, refine features, integrate with systems with frequent stakeholder check-ins

Handover and deployment - deliver AI solution into your systems

Knowledge transfer with documentation and training of internal engineers

Post-deployment monitoring and fine-tuning

Processing Time

Quarterly project intakes, 3-6 month project duration depending on variant selected

Project Duration & Completion

6 months for MVP path or 3 months for PoC variant, structured in phases: assessment/scoping → development → handover/knowledge transfer

Best Practices for a Successful Application

Ensure data readiness - sufficient, high-quality, structured data is critical for model performance

Set realistic scope - avoid trying to do too much in the timeframe to prevent incomplete or fragile solutions

Engage internal team actively - ensure your staff absorbs knowledge for post-project maintainability

Plan for post-deployment sustainability - consider scaling, maintenance, integration, versioning, monitoring resources

Understand IP/ownership arrangements - rights may be shared or contested depending on cash/in-kind contribution split

Implement governance from start - ensure responsible AI design and auditing according to standards

Prepare strong proposal - selection is competitive, so case and plan must be comprehensive

Address bias and compliance risks - AI systems must meet responsible AI standards

Focus on real business problems - avoid generic challenges that have existing off-the-shelf solutions

Commit to knowledge transfer - maximise learning opportunity for internal capability building

Sample Use Cases & Scenarios

Sompo Holdings: Built AI solution for fraudulent claims detection, improving detection rates and automating lower-risk claims

IBM: Used 100E to classify product quality/defect risk via AI models, reducing human review time from a week to an hour

EM2AI (healthcare): Developed AI for automated dental charting using imaging, deployed to clinics

Insurance fraud detection systems

Medical analysis and diagnostic tools

Logistics routing optimisation

Real-world business problem transformation into AI solutions

Additional Information

Project Structure and Variants: 100E (3-month POC): • Shortened, intensive path to validate AI concept quickly • Same methodology and team with AI Singapore's engineers + apprentices • Ideal for organisations wanting faster test before scaling • Delivers validated proof-of-concept 100E (6-month MVP/standard): • More time to build, refine, and deliver production-ready AI solution • Greater depth, possibly more features, more integration • Full production-ready deliverable Key Strengths and Advantages: • Access to AI expertise and talent - leverage AI Singapore's engineering team, domain experts, and AIAP apprentices • Co-funding reduces cost burden - grant support and matching model lower financial barrier to AI experimentation • Capability building - internal staff learn from process and absorb skills • Structured methodology and governance - built on disciplined processes and Model AI Governance Framework • Proven track record with credible projects across multiple industries Intellectual Property Arrangements: • IP arrangements differ depending on variant (industry vs research) • Often jointly owned or fully by sponsor depending on cash/in-kind split • Clear agreements established during project scoping phase Risks and Challenges: • Data limitations can derail model performance • Overambitious scope can lead to incomplete solutions • Dependence on external team if internal team doesn't engage • Selection competitiveness - strong proposals required • Governance, bias, and compliance risks must be managed