What separates a useful healthcare chatbot from a generic one?\u00a0\u00a0<\/p>\n
\u00a0<\/p>\n
For the winning\u00a0team at the recent NUS-Synapxe-IMDA AI Innovation challenge, the answer lies in the\u00a0foundational models underneath.\u00a0\u00a0<\/p>\n
\u00a0<\/p>\n
Team Assure, which compromises of six students from National University of Singapore (NUS)\u2019 Business Analytics course, built a voice-enabled artificial intelligence (AI)\u00a0companion within four months\u00a0that\u00a0conducts daily check-ins with elderly cardiac patients.\u00a0<\/p>\n
\u00a0<\/p>\n
The check-ins were done in\u00a0ways that reflect local language patterns, cultural\u00a0nuances, and the sensitivity that care settings demand.\u00a0<\/p>\n
\u00a0<\/p>\n
To do so, the team tapped into Singapore\u2019s homegrown\u00a0models,\u00a0SEA-LION and\u00a0MERaLiON, to\u00a0\u201cmove beyond a generic chatbot and design something natural and familiar for elderly users,\u00a0without compromising on the human-in-the-loop safeguards that sensitive care scenarios demand,\u201d says a spokesperson from Team ASSURE to\u00a0GovInsider.\u00a0<\/em><\/p>\n \u00a0<\/p>\n While\u00a0SEA-LION (Southeast Asian Languages in One Network)<\/a>\u00a0focuses on text and data,\u00a0MERaLiON (Multimodal Empathetic Reasoning and Learning in One Network)<\/a>\u00a0goes beyond standard text-based AI by integrating multiple modes like speech-to-text, emotional and paralinguistic intelligence and more.\u00a0<\/p>\n \u00a0<\/p>\n For the team, using\u00a0local models\u00a0was not an afterthought, but what made the solution\u00a0viable\u00a0for the people it was built to serve.\u00a0\u00a0<\/p>\n \u00a0<\/p>\n \u201cHaving LLMs that genuinely understand local language patterns and cultural\u00a0nuance\u00a0is what makes that possible,\u201d he says.\u00a0<\/p>\n \u00a0<\/p>\n The winning team clinched\u00a0the\u00a0S$10,000 top prize, which was announced at the conclusion of the challenge on April 22.\u00a0<\/p>\n \u00a0<\/p>\n In its 12th\u00a0edition, the challenge\u00a0was\u00a0organised\u00a0by NUS,\u00a0Singapore\u2019s\u00a0Infocomm\u00a0Media Development Authority (IMDA) and\u00a0the\u00a0national\u00a0healthtech\u00a0agency Synapxe.\u00a0\u00a0<\/p>\n \u00a0<\/p>\n This edition\u00a0welcomed 880\u00a0students from 18 institutes, including universities,\u00a0polytechnics\u00a0and junior colleges, forming 181 teams to develop AI-powered solutions for chronic conditions\u00a0like diabetes,\u00a0hypertension\u00a0and high lipid levels.\u00a0<\/p>\n \u00a0<\/p>\n Most students used the SEA-LION and\u00a0MERaLiON\u00a0models, including the top eight teams, to develop solutions for this challenge.\u00a0\u00a0<\/p>\n \u00a0<\/p>\n IMDA’s involvement in the challenge reflects broader ambition: To\u00a0translate the use of SEA-LION and\u00a0MERaLiON\u00a0models into solutions that serve Singapore and the region.\u00a0<\/p>\n \u00a0<\/p>\n Its Director for Business & Ecosystems at its\u00a0BizTech\u00a0Group, Dr\u00a0Lawrence\u00a0Wee,\u00a0said the agency’s goal was to cultivate an ecosystem of partners and innovators to enable that.\u00a0<\/p>\n \u00a0<\/p>\n Developed under the\u00a0National Multimodal Large Language Model\u00a0Programme\u00a0(NMLP)<\/a>,\u00a0both\u00a0models\u00a0were\u00a0built to handle the code-switching, mixed-language communication styles, and cultural nuances\u00a0common in the local Singaporean setting.\u00a0<\/p>\n \u00a0<\/p>\n In healthcare particularly,\u00a0Dr Wee\u00a0noted that such models can help bridge the language gap<\/p>\n between elderly patients and care providers, enabling care that is not just efficient but more empathetic and inclusive.\u00a0<\/p>\n \u00a0<\/p>\n The NMLP Special Award\u00a0of S$5,000\u00a0went to\u00a0VitalSense, built by Team Wait for a Name, which generates a health summary in five minutes using a webcam, microphone, and self-reported vitals\u00a0\u2013 all\u00a0without a clinic visit.\u00a0<\/p>\n \u00a0<\/p>\n \u201cSEA-LION didn’t just power our reporting\u00a0layer,\u00a0it raised the bar for what an AI-generated health summary could look like in a real-world clinical context,\u201d a spokesperson from\u00a0the team\u00a0tells\u00a0GovInsider.\u00a0<\/em><\/p>\n \u00a0<\/p>\n The team used SEA-LION to\u00a0synthesise\u00a0multimodal inputs –\u00a0camera footage, motion data, voice, and questionnaire responses –\u00a0into\u00a0health reports.\u00a0\u00a0<\/p>\n \u00a0<\/p>\n It was also able to generate\u00a0structured summaries, risk interpretation,\u00a0and recommendations, he adds.\u00a0 \u201cTo\u00a0maximise\u00a0reliability, we\u00a0standardised\u00a0input and output formats, while building a backend fallback mechanism that allows the system to deliver consistent results even during large-text processing, extended outputs, or occasional API instability,\u201d he\u00a0explains.\u00a0<\/p>\n \u00a0<\/p>\n Synapxe\u2019s\u00a0Assistant Chief Executive Alan Goh tells\u00a0GovInsider<\/em>\u00a0that the winning use cases will be shared with public healthcare clusters, and that the agency is working with NUS to extend internship opportunities to selected participants who want to pursue healthcare AI capstone projects.\u00a0\u00a0<\/p>\n \u00a0<\/p>\n \u201cThe goal is to build skills and insights essential to refining and strengthening ideas, rather than future implementation,\u201d he shares,\u00a0adding that the challenge\u00a0provides\u00a0an opportunity for young innovators to immerse in real-world healthcare challenges and constraints.\u00a0<\/p>\n \u00a0<\/p>\n The intention is to\u00a0cultivate an innovation\u00a0and\u00a0design thinking\u00a0mindset among\u00a0students.<\/p>\n \u00a0<\/p>\n For public healthcare partners, challenges like this\u00a0also\u00a0offer an early window into\u00a0the emerging\u00a0ideas,\u00a0approaches\u00a0and capabilities.\u00a0<\/p>\n \u00a0<\/p>\n Goh is candid that the path from\u00a0a\u00a0challenge\u00a0entry\u00a0to real-world deployment\u00a0depends\u00a0on multiple factors: real-world demand, ability to integrate with existing workflows and infrastructure, and rigorous assessment by multiple stakeholders.\u00a0<\/p>\n \u00a0<\/p>\n Getting the path\u00a0right, he adds, is\u00a0ultimately what\u00a0separates\u00a0a\u00a0good idea\u00a0from a practical solution\u00a0that frontline teams can\u00a0rely\u00a0on.\u00a0<\/p>\n \u00a0<\/p>\n Second prize of S$7,000 went to Team\u00a0SilverGait\u00a0for an AI system that monitors elderly adults for frailty using computer vision and wearable data, while the third place of S$5,000 went to Team Med-SEAL for a diagnostic platform that supports clinicians by\u00a0automating\u00a0clinical reporting and delivering\u00a0clearer, multilingual medical explanations to patients.\u00a0<\/p>\n \u00a0<\/p>\n What separates a useful healthcare chatbot from a generic one?\u00a0\u00a0 \u00a0 For the winning\u00a0team at the recent NUS-Synapxe-IMDA AI Innovation challenge, the answer lies in…<\/p>\n","protected":false},"author":1,"featured_media":1864,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"","fifu_image_alt":"","footnotes":""},"categories":[2611],"tags":[],"class_list":["post-56168","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-buzz-headlines","wpcat-2611-id"],"_links":{"self":[{"href":"https:\/\/sgbuzz.com\/index.php?rest_route=\/wp\/v2\/posts\/56168","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/sgbuzz.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/sgbuzz.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/sgbuzz.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/sgbuzz.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=56168"}],"version-history":[{"count":0,"href":"https:\/\/sgbuzz.com\/index.php?rest_route=\/wp\/v2\/posts\/56168\/revisions"}],"wp:attachment":[{"href":"https:\/\/sgbuzz.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=56168"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/sgbuzz.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=56168"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/sgbuzz.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=56168"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}Homegrown LLMs finding their way into the home\u00a0<\/h3>\n
![\"\"](\"https:\/\/gov-web-sing.s3.ap-southeast-1.amazonaws.com\/uploads\/2026\/5\/image-1779969039524.jpg\")
\u00a0<\/p>\nNext steps following the challenge\u00a0<\/h3>\n
![\"\"](\"https:\/\/gov-web-sing.s3.ap-southeast-1.amazonaws.com\/uploads\/2026\/5\/image-1779763597686.jpg\")
<\/a>
\n
\n
Read Full Article At Source <\/a>
\n