Fact-Checking AI to Cancer Data: Three Singapore Startups Take Homegrown Tech Global
Source: The Business Times
As Singapore turns 61, The Business Times turned the spotlight on three startups at JTC's LaunchPad @ one-north that are chasing global problems with homegrown technology — a useful snapshot of how far the Republic's deep-tech scene has come from its consumer-app roots. All three sit in the same building,...
As Singapore turns 61, The Business Times turned the spotlight on three startups at JTC's LaunchPad @ one-north that are chasing global problems with homegrown technology — a useful snapshot of how far the Republic's deep-tech scene has come from its consumer-app roots. All three sit in the same building, yet each is attacking a different frontier: machine-speed fact-checking, cancer tissue analysis and VR-based rehabilitation.
AI Seer, founded in 2019 by former due-diligence professional Dennis Yap, builds "truth-seeking AI." Its flagship Facticity.AI platform cross-references claims against multiple sources and returns the references behind its conclusions — during a 2024 US presidential debate it checked almost 250 claims in near real time, earning a spot on Time's best inventions list. The company also sells a contactless polygraph called the Multi-Spectral Reality Detector, developed with NUS and A*Star researchers, and its newest product, ArAIstotle, verifies claims in real time as they are made.
ImmunoQs, spun off from A*Star in 2022, tackles a data problem: cancer tissue images that can run to 200 gigabytes per file, too big for most hospitals to analyse. Its platform visualises and analyses spatial tissue data to map the tumour microenvironment, helping clinicians identify which patients are likely to benefit from immunotherapy — relevant since only around 20 per cent of patients respond to the treatment, with bills running as high as S$300,000 for those who don't. Revenue doubled to about S$580,000 in 2025, the company is profitable without external funding, and its customers include Singapore General Hospital and the National Cancer Centre.
The third, DancingMind, turns rehabilitation into virtual reality games. Founded by Jennifer Zhang in 2017, the healthtech startup's exercises — from tidying a messy virtual living room to guiding a pet through a maze — train memory, attention and decision-making for dementia and stroke patients, with the system tailoring programmes to each user's progress. It now has paid contracts with more than 70 per cent of Singapore's restructured hospitals.
What links the three is a deliberate bet on data and trust rather than models and compute. ImmunoQs' Joey Lee puts it plainly: "Data is really the foundation for building AI — usable, structured data." AI Seer's Yap argues generative AI has collapsed the cost of producing convincing falsehoods, so verification itself must run at machine speed. Both see Asia-specific data — cancers more prevalent in Asian populations, clinical norms that differ from the West — as an edge no Silicon Valley lab can easily replicate.
Why it matters for Singapore: These are exactly the companies Singapore's AI strategy was designed to produce: deep-tech startups that use the Republic's research base, subsidised infrastructure and trust positioning to sell into global markets. JTC's refreshed LaunchPad master plan, including the new Kampong AI community, is doubling down on this model. The National Day feature is a reminder that Singapore's AI story is no longer just about attracting foreign giants — it is increasingly about exporting its own.
