“Want to see an orangutan or white tigers first?” A female avatar shines from a large screen attached to the wheels. If you only have a few hours to spare when visiting the Singapore Zoo, Mandy, a concierge robot, can help you design your instant zoo itinerary.
Mandy is one of the autonomous mobile robots that has been trialled by the Mandai Wildlife Group. She can help you with directions, provide fun facts about zoo animals, or simply lead you to where you want to go. If dark clouds gather, she pulls out live weather data and proposes a rainy weather plan instead.
Her intelligence is powered by it
IBM Data and Artificial Intelligence (AI) Platform known as Watsonx
. Trained with Mandai Wildlife Group data, Mandy can personalize responses to the needs of any visitor and even respond to facial expressions.
IBM's cutting-edge generation AI solutions provide the responsive and flexible customer experiences that Mandai Wildlife Group needs. “We've seen an important role in the development of our services,” said Ang Ai Kiar, IBM Data Scientist. “We leverage our own generative AI capabilities to show how Mandy can create new responses on the fly, rather than being limited to pre-programmed answers.
Concierges with Gen AI may seem easy to implement, but turning AI ambitions into practical solutions is a real challenge for most companies.
Large-scale language models (LLMs) may have taken the world into a storm, but they are not always business-friendly. For example, it may come with high costs, security concerns, and limited customization options regarding your specific business needs.
Catherine Lian, general manager and technology leader at IBM ASEAN, said smaller, industry-class AI models can offer significant benefits to businesses.
Beyond the resource challenges, businesses are concerned about the tendency for data security and Gen AI to compensate for “hastique” or false answers.
Despite these toothaches, businesses are seeing real productivity gains from Gen AI. According to an IBM survey,
57% of Singapore organizations plan to increase their AI investment in 2025
. The biggest investments are in areas such as IT operations, data quality control, and finance operations.
To address these concerns about resource requirements, data security and AI accuracy, IBM is heading a different path than other AI companies. This highlights openness, transparency and cost-effectiveness.
First of all, IBM's Watsonx platform uses an open source AI model known as the Small Language Model (SLM), which can save up to 98.5% on costs. Unlike closed-source AI models that behave as “black boxes” with their own code and training data, training data that users cannot look up offers much more transparency in open-source SLM.
Another application for IBM's small language model of granite is the Ferrari Scuderia HP mobile app. This translates complex race data into fan-friendly insights such as overtaking analysis.
Photo: IBM
Organizations can also inspect their code, understand how the model works, and modify it to suit their business needs. This open source approach allows businesses to leverage the Watsonx platform for enterprise-grade hosting, updates and support services while maintaining control and transparency over their AI models.
Open source models become more accurate and reliable over time, as global community developers can discover mistakes, fix technical issues, and add new features. For businesses, this means getting new features faster. They don't stick to one vendor's technology and can customize AI capabilities for their customers.
According to a survey of more than 2,400 IT decision makers from 21 countries,
51% of companies using open source AI tools report positive returns (ROI) of their investment
Compare with only 41% of people using a closed source system.
Expanding our commitment to open source AI,
IBM has released the latest version of the Granite Model Family, Granite 3.2.
February 2025. Granite is IBM's SLM collection that powers the Watsonx platform.
The new Granite 3.2 model can match or exceed the performance of larger models while using less power. These include specialized models of various tasks for financial forecasts up to two years, supply chain planning and inventory management. Companies can access these granite models and integrate them into other popular models such as Meta's llamas and Mistrals via IBM's watsonx.ai platform.
Granite's strengths shine in environments like the Singapore Zoo and other wildlife parks. Robots equipped with granite SLM do not require a certain internet connection. It can operate offline while using minimal power. This allows the entire compound to serve visitors for a long period of time. Their flexibility also allows them to perform specific tasks such as tracking animal movements, identifying species, and monitoring environmental conditions.
Ang said: “Granite SLM offers speed, cost efficiency, and adaptability, making it ideal for continuous operation in resource-constrained environments such as wildlife sanctuaries.”
Granite SLM is also turned on
New Scuderia Ferrari HP Mobile App brings F1 fans closer to their favorite teams than ever before
. Fans can read the race summary within hours of the race conclusion. Dynamic visuals created using AI technology in Watsonx interact with post-race data, including telemetry, weather, track conditions, session results, and car and tyre strategies.
Cost comparison:
-
Cost 98.5% lower than closed source AI models
-
51% of open source users report positive ROIs (41% for closed models)
Technical Benefits:
-
Requires less computing power and reduces operational costs while matching the performance of larger models
-
Reduce development costs through community-driven improvements rather than independent development
Business Flexibility:
-
Eliminate vendor lock-in costs and allow for more flexible and economical scaling
IBM's open source approach offers many technical and cost benefits, but the company recognizes that technology alone is not enough. As Gen AI becomes more powerful, businesses are increasingly concerned about trust, safety and management. While some AI companies are shrinking their ethics teams, IBM has doubled their responsible AI.
This approach to responsible AI is not only a technical consideration for IBM, but also central to our vision of how AI should be developed and deployed.
Lian said: “At IBM, we have a purpose to bring groundbreaking innovation to AI. AI is a deep opportunity, but it has a high interest.
“Imagine your organization's generative AI compromises your customer's personal data or scams them. This simply can't happen.”
As Gen AI disrupts and augments businesses, AI-driven assistants like Mandy at Mandai Wildlife Group show that businesses can already benefit from using IBM's agile and ethical framework.
Discover how IBM's trustworthy AI framework can transform your business today.
