Finding a job in Singapore as a Malaysian is genuinely hard. A role needs to fit the work, clear a salary threshold and realistically support an Employment Pass. Reading every listing against a real profile takes time I did not have, so I built a multi-agent system to do the searching and preparation with me.
Evidence before recommendation
The system imports a résumé, portfolio, certificates and chosen career material only after preview and consent. It separates confirmed facts, unconfirmed claims and goals instead of silently merging them. Conflicts return to the user.
For every role, it explains what I satisfy, what I do not and what to strengthen. Each conclusion links back to evidence. Visa and salary constraints are first-class filters, not disclaimers added after a generic “good fit” score.
Pressure-test the résumé
The most useful interaction checks each line and asks: could I defend this in an interview? Challengeable wording is flagged. A tailored version can change emphasis, but it cannot invent proof.
Discovery can run on a schedule through approved job sources and push fresh roles. Applying remains a human action. The automation removes searching and preparation work without taking control of consequential decisions.
A useful career agent should make a candidate more truthful and prepared, not merely more prolific.