Singapore's VisionX AI Launches AEGIS Graph Model to Map Crypto Crime Networks
Key Takeaways
- •AEGIS, short for Attention-based Entity and Graph Intelligence System, is a Large Graph Model pretrained on financial transaction graph data to map how wallets, accounts and entities connect and to reveal patterns linking seemingly unrelated activity.
- •TraceX, the first product built on AEGIS, targets financial crime investigators and law enforcement with planned capabilities including reverse tracing, MegaMap network analysis, pattern recognition and case association.
- •The TraceX platform keeps final decisions with human investigators, recording each decision and backing leads with traceable evidence so accountability rests with people rather than automated systems.
- •VisionX AI's wider suite includes VisionX KYT, which aggregates blockchain risk data covering 1 billion wallets across 25 chains, and Web2.5 TxM, which monitors transactions across both traditional finance and digital assets.
- •The launch follows a July 2026 warning from the Financial Action Task Force that criminals use stablecoins, cross-chain bridges, decentralized exchanges and mixers to hide illicit funds.

SINGAPORE — VisionX AI, a Singapore-based company within Alpha Ladder Group, has launched AEGIS, a proprietary artificial intelligence model built to detect financial crime networks. The model is designed to uncover links between accounts and crypto wallets that may appear unrelated at first glance. Alongside the model, the company introduced TraceX, a blockchain investigation platform and the first product built on AEGIS.
The launch took place at an industry forum held by MetaComp in Singapore on November 17. VisionX AI was first developed inside MetaComp, the group's payments business, and now operates as a standalone regulatory technology company. News of the launch was shared by the Press Releases newswire account on X on November 18.
How the AEGIS Model Works
AEGIS — short for Attention-based Entity and Graph Intelligence System — is a Large Graph Model built on a Graph Transformer design. It was pretrained on financial transaction graph data to learn how wallets, accounts and entities connect with one another. In this kind of model, addresses and entities become points in a network and the flows between them become connections, so analysis can follow value across an entire system rather than examining each transfer in isolation. The name draws on Greek mythology, where the aegis was a shield carried by Zeus and lent to Athena.
Rather than operating on its own, AEGIS is paired with large language models from established providers in a hybrid setup. The graph model handles network analysis, while the language models handle reasoning, plain-language questions and investigation workflows.
Tin Pei Ling, Executive Chairwoman of VisionX AI, said financial crime runs through networks of people, entities and financial activity. She said AEGIS learns how money moves between accounts and wallets, allowing it to spot patterns that link activity which may otherwise look unrelated.
According to the company, tools that review one transaction at a time can struggle to see the larger network. The launch comes amid heightened regulatory attention on the issue: in July 2026, the Financial Action Task Force — the intergovernmental body that sets global standards for anti-money laundering — warned that criminals use stablecoins, cross-chain bridges, decentralized exchanges and mixers — services that pool funds from many users to obscure where the money originated — to hide illicit funds.
Singapore's VisionX AI unveils its own hybrid AI model to uncover financial crime networks across accounts and wallets At its core is AEGIS, a Large Graph Model built for financial crime, which will power TraceX for on-chain investigations SINGAPORE,… — Press Releases (@press_newswire) November 18, 2026
TraceX Puts Investigators in Charge
TraceX is being built for financial crime investigators and law enforcement agencies. The platform is intended to trace fund flows across blockchain addresses and uncover hidden links between seemingly separate cases. Planned features include reverse tracing, MegaMap network analysis, pattern recognition and case association — capabilities meant to connect individual cases to the larger criminal groups behind them.
Human oversight remains central to the design. Investigators will make the final decisions, guided by the platform's suggestions. Each decision will be recorded, and leads will be backed by traceable evidence and human review. The structure is designed to keep accountability with human investigators rather than with automated systems.
Summer Yu, Chief Executive Officer of VisionX AI, said individual cases often look unrelated even when the same criminal network is behind them. She said AI shows investigators where to look next, but their judgment and the evidence still decide the case.
The Broader VisionX Platform
AEGIS and TraceX sit within VisionX AI's wider product suite. VisionX KYT, a Know-Your-Transaction service, pulls blockchain risk data from several analytics vendors, covering 1 billion wallets across 25 chains. The platform also offers Web2.5 TxM, which monitors transactions across both traditional finance and digital assets. VisionX won the Best Compliance-First Innovation award for transaction monitoring at The Digital Banker's Digital Assets Awards 2026.
VisionX also supports compliance for MetaComp's StableX Agentic Network, which launched as the StableX Network in November 2025. That network now has more than 1,000 members reaching over 100,000 users. MetaComp is licensed by the Monetary Authority of Singapore as a Major Payment Institution, a licensing category under Singapore's Payment Services Act that covers payment services above regulatory thresholds. Its network lets AI agents access regulated payment services, supported by a governance tool called StableX KYA that launched in April 2026.
Expansion as a Standalone Company
As a standalone company, VisionX AI plans to offer its products to financial institutions, digital asset firms and public-sector groups. It also aims to expand its graph technology into fraud and scam prevention, compliance risk management and investigations. Near-term developments to watch include the rollout of TraceX's planned features and the response from the financial institutions, digital asset firms and public-sector groups VisionX AI plans to serve.