Accelerating AI with 100E
Navigating AI adoption can be daunting. To inspire action, we'll showcase three real-world examples where strategic implementations delivered immense value. From increased revenue to improved efficiency and customer satisfaction, these proven use cases demonstrate the transformative power of AI across business functions. By highlighting these 100E successes, we aim to help leaders recognize the immense opportunities within reach. Let these three inspiring examples ignite ideas for how AI could soon improve your own organization.
About the 100E
The 100 Experiments (100E) programme by AI Singapore's (AISG) Innovation pillar aims to catalyze AI innovation and adoption in Singapore's economy. It provides a platform for companies to collaborate with research institutes on applied AI projects. Through rapid prototyping over 7 months leading to a Minimum-Viable-Model (or Product), the goal is to quickly develop impactful AI solutions for industry-relevant challenges. 100E allows companies to experience hands-on what AI can achieve for their business. The programme has supported over 100 projects across various sectors since its launch in 2018.
In addition, the 100E is also the platform which we use to train the apprentices undergoing the AI Apprenticeship Programme (AIAP). This was described in the article: The AIAP (and 100E) Experience - AI Singapore Community
Case 1: Sompo's AI-Powered Fraudulent Claims Detection System
Singapore-based insurer Sompo Holdings (Asia) participated in the 100E programme for its AI-powered fraudulent claims detection system in 2019.
Together with AISG's AI Engineers and Apprentices, Sompo built a machine-learning model over 7 months that could flag suspicious claims for further investigation. The model was trained on thousands of past travel, and personal accident claims data provided by Sompo.
The new AI system provides 100% fraud detection coverage and has improved detection rates by 90%. Customers with approved claims now receive payouts much faster - potentially within minutes.
Sompo also benefits from significant cost savings from reduced manual checking, estimated at $200k-300k annually by year 3 of deployment. Employees also spend less time on manual reviews and can focus on investigating complex suspicious cases.
After successful pilots in Singapore and Indonesia, Sompo plans to expand AI fraud detection to other markets and lines of business like motor insurance.
The project showcases how insurers can use AI to improve efficiency and customer experience. Collaboration through programmes like 100E also allows companies to develop impactful AI solutions quickly. Sompo's win at the recent awards is a testament to the success of the 100E programme.
The project also won the AI Award for general insurance from Singapore Business Review in 2020.
Case 2: IBM Improve Product Quality with AI
IBM collaborated with AI Singapore (AISG) on a 100 Experiments (100E) project to improve product quality classification for its mainframe line using AI in 2019.
IBM wanted to help its product quality engineers better analyze performance data to identify issues early.
Previously, engineers manually analysed numerous charts and tables to spot anomalies - a tedious and time-consuming process. Collaborating closely with the IBM team, the 100E team built a deep learning model to categorize product risk levels based on performance graphs images. The model was trained on 10 years of historical data.
The model mimics how engineers visually identify anomalies. It achieved over 90% accuracy in classifying products as high, medium or low-risk. High-risk products are flagged to engineers for investigation. This enables engineers to focus on investigating root causes rather than manual monitoring. Importantly, it has also reduced the time taken from 1 week to 1 hour.
The collaboration demonstrates how the 100E programme enables companies to develop impactful AI solutions quickly by tapping AISG's expertise. The improved product quality engineering will ultimately benefit IBM's customers who rely on its mainframe systems.
Case 3: EM2AI Develops AI for Automated Dental Charting in 100E Collaboration
Local healthcare startup EM2AI (a deep-tech spin-off from Q&M Dental) collaborated with AI Singapore (AISG) on a 100 Experiments (100E) project to create an AI solution for automated dental charting in 2019.
Dental charting during examination is a tedious and manual process. EM2AI, in collaboration with AI Singapore, developed a computer vision model that can automatically detect pathology from dental X-rays and extract key information like tooth numbering, caries, and missing teeth.
AI Singapore undertook the 100E 7-month project with AISG apprentices who helped build and refine the AI model. The AI model aids dentists by reducing time spent on manual charting and minimizing errors. By standardizing dental records and diagnosis, the AI tool can also improve the quality of care.
As of 2023, the solution has been rolled out to over 150 dental clinics in Singapore and Malaysia.
These projects across insurance, manufacturing, and healthcare show how companies can leverage AI to enhance productivity, decision-making, and service quality. The collaborations between AI Singapore and companies like Sompo, IBM, and EM2AI highlight the success of the 100E program in driving industry AI adoption. 100E enables rapid prototyping of impactful AI solutions tailored to business needs by catalysing applied research and local AI talents.
In addition, by leveraging 100E as a real-world training platform for apprentices, we have trained more than 300 Singaporeans as AI Engineers. This holistic win-win-win approach for AI Singapore, the company and the apprentices is a sustainable and cost-effective methodology to train AI Ready AI talents for the industry.
Outcompute to outcompete | Growing our own timber