This is a game simulation case on Machine Learning (ML) and can be used in the classroom to introduce to students the foundational concepts of ML. The case begins with a brief background on ML, its origins, the objectives of its usage across various industries, and ends with a future outlook on the relevance of ML. At the end of the case, a link to the game simulation is provided, which can be played by students in the classroom. The main topics covered in the game simulation include: • An introduction to ML • Different categories of ML algorithms including unsupervised, supervised and re-enforcement learning (based on the differences in the way they process the data for decision-making). • Identifying applications and use cases of ML

Set in April 2020, the case talks about TAIGER, a software-as-a-service (SaaS) company providing natural language processing (NLP) solutions in a rapidly growing market where demand and competition for such solutions are high. TAIGER solutions could process and digitise large amounts of physical data, perform search and extract functions on data, and make appropriate recommendations. The algorithms and tools were bundled under three packages, and were customised for every client. Client companies were mainly large organisations and government entities. Despite a growing customer base, Arroyo and his team found it increasingly difficult to service new clients, who demanded more customisation and services. TAIGER's solution packages were bundled together with customisation and post-implementation support services based on contract licenses. The downside of this model was that it used many resources and limited the delivery of the products to a per project basis. The monetisation of the model was also complicated and project costs were difficult to control. Administering customised solutions was time consuming and expensive for both TAIGER and its clients; it also lacked flexibility and quick scalability for large-scale implementation. Arroyo realised that he needed more than just efficient solutions, given the expanding opportunities for NLP in the market and the constraints of TAIGER's existing solution packaging. He wondered if designing a new business model was the right way forward. Would he also need to devise a new pricing strategy and rebundle solution offerings? How could he unbundle the business value of TAIGER's NLP solutions?

Set in April 2020, with the backdrop of the Covid 19 pandemic, the case talks about the opportunities for visual search in the online retail market segment and beyond. Visense is a visual search software-as-a-service (SaaS) solutions provider with a pay-as-you-use, API based, market solution that focuses on the retail segment. Visenze had experienced a spike of about 30% in the usage of its solutions between January to April 2020. Predominantly, usage growth had extended across all item categories, with footwear, apparel and jewellery products being the top gainers. The presumed understanding from this trend was that with more time in hand for home browsing during the pandemic, consumers were exposed to more visual inspirations and objects of desire. This inclination had led them to explore ecommerce websites increasingly, paving a direct path to progressive sales conversions. Moreover, consumers had started to demand shorter purchase journeys, in the quest for a frictionless ecommerce experience. While Generation Y and Generation Z were the major users of ecommerce and visual search, older generations (Generation X and Baby Boomers) had also started to increasingly use online channels of shopping. Visual search was firmly in a sweet spot with the promise of collapsing the conversion process from image to purchase in a few seconds. Amidst such market conditions, Oliver Tan, co-founder of Visenze, wondered if his firm required a new market strategy to tap on the latest opportunities. Should Visenze continue to focus on the retail segment? Should it target older consumers in the retail segment? Will its technology need to be enhanced to target the new consumer segment?

In March 2020, Manav Kamboj, CTO of PropertyGuru, an online property platform, wondered what it would take for his firm to become a one-stop solution-shop for property seekers in the home buying process. His company had just launched a new application - PropertyGuru Finance which could enable seekers to search for a home loan entirely online. The solution had taken Kamboj and his team one step closer to achieving the organisation's vision - enabling completely online property transactions by 2025. PropertyGuru had implemented several AI-based solutions to transform its platform from a property-search to a property-trust platform. The goal was to become the customer's trusted friend, philosopher and guide in the home buying process. The transformation had seen the implementation of several useful analytics tools and AI applications that had helped property seekers, sellers and agents make informed decisions in buying and selling properties. It had also propelled the company to become Asia's leading online property platform providing users a choice of over 2.7 million homes and the preferred destination of more than 24 million property seekers per month. To attain its vision of enabling fully digital property transactions, however, the company would need to think outside the box and rely on its old friend - technology, to design new solutions to enable online property transactions. How could PropertyGuru become an end-to-end property solution platform? What emerging trends in real estate could the company explore? Would it need to bundle all its proposed applications, or could it offer the additional offerings as standalone services? Could AI open more doors for the company?