What we do
Three domains, built one at a time until the work is measurably better.
We build around industries, not features. We pick a field, learn how the work is done, and build agents that hold up in it. Three fields are furthest along today.
Domain 01
Cybersecurity
A security operations centre produces more alerts in a day than its analysts can read in a week. The usual response is to add more tooling, which produces more alerts. Agentic AI only helps here if the team trusts it to act, and in a SOC trust means bounded, logged and open to review.
Our agents take on the parts of the job that do not scale with headcount. They read the alert, gather the context an analyst would have gathered, form a view, and hand the analyst a decision instead of a queue.
Triage
Alerts assessed, enriched and ranked before an analyst opens them. False positives are closed with the reasoning recorded, never silently suppressed.
Adversary simulation
Safe, continuous emulation of attacker behaviour, so a control that has drifted is caught in the same week, not at the annual test.
Monitoring and analysis
Sustained watch across signals and sources, with analysis written for the person who has to make the call.
Response support
Timelines assembled, containment options drafted and reporting written while an incident is still live, with every action attributable.
Securing the agents themselves
Defending an organisation with agents raises an obvious question: who defends the agents. The platform answers it with a validation agent that tests what is built on it, within a scope the customer sets, and reports findings with the evidence behind them.
Domain 02
Education
From the 2027 academic year, students across Singapore's institutes of higher learning are expected to receive compulsory AI skills training. That turns a procurement question into a curriculum one: what do students build, who supervises it, and what does the institution learn from it.
We provide a learning environment, not a software licence. Students build agents for the discipline they are studying. Lecturers work in the same space and shape the project as it develops. What the cohort produces feeds back to the people teaching it.
The analytics are built for differentiated instruction: seeing which students are keeping up, and directing classroom time where it will do the most good.
Built in their own discipline
Students across faculties build agents for their own field, whether that is fintech, manufacturing or marketing communications. The AI work is the coursework, not a separate module bolted on beside it.
Lecturers in the same space
Educators work alongside the cohort in the same environment, shaping a project while it is still taking form instead of marking it once it is finished.
Signals a teacher can act on
Classroom recordings and academic performance data become insight for the educator, grounded in established pedagogical practice instead of left as a dashboard to interpret.
The institution keeps it
What a cohort creates stays with the institution, retained and retrievable long after the students graduate, along with the record of how it was used.
Domain 03
Fintech
Capital moves on judgement made under time pressure, and every automated step behind that judgement has to be explainable to a regulator afterwards. Those are the conditions the platform was built for.
The work splits two ways: helping people read a situation before they commit to it, and carrying the review burden behind every client and every transaction.
Decision intelligence
Investment cases and sales performance taken apart line by line, so an investor or an operator gets a clearer read on the numbers before capital is committed.
AML and KYC
Anti-money-laundering alerts and know-your-customer checks assessed, enriched and ranked before a reviewer opens them, with the reasoning recorded so the decision can be defended later.
Entity research
Research that maps who sits behind a counterparty and the risk that travels with them, assembled from the sources an analyst would otherwise work through by hand.
Data that stays put
Deployment inside the institution’s own environment, with customer-held keys, for firms whose data cannot leave the jurisdiction.
How we go to market
We start from relationships that already exist.
We lead commercialisation across the platform: finding the opportunities, taking the domains to market and holding the customer relationship.
Start from the installed base
Our parent company already sells into enterprises across the region, so most conversations start from an existing relationship.
Subscription, then use
A recurring subscription for the domain, plus consumption charges for the work the agents perform. Easy to budget for, and clear about what usage costs.
Prove here, then expand
Singapore is our reference market. We prove a domain here, then take it across the region on the same platform, localised for each jurisdiction.
Who we serve
Enterprise security teams, regulated organisations with data-residency obligations, institutes of higher learning, and public bodies that need to demonstrate sovereignty.
Capability, security and deployment options vary by domain and release stage, and are set out in your agreement with us.