
Helios: the review happens before the signature
Colombia's Defense Sector brought data analytics and artificial intelligence into the oversight of its own procurement. The system cross-references public contracting platforms and raises alerts while a tender is still open.
Corruption in public contracting is rarely caught in time. It surfaces once the contract has been executed, once the money is gone, and once the only remaining recourse is a disciplinary process that will take years. The diagnosis behind the Helios System starts there: the problem was never a shortage of information — Colombian public procurement is among the most thoroughly documented in the region — but the human impossibility of reading it fast enough.
What the system does
Helios cross-references SECOP I and SECOP II, the two platforms on which the Colombian state publishes its contracting processes, and applies data analytics and artificial intelligence to that cross-reference in order to identify red flags and detect unusual patterns. Three signals draw most of its attention: tenders that close with a single bidder, participation by individuals or companies under active sanction, and unusual concentrations of awards in the hands of one actor. None of the three, on its own, proves an irregularity. All three, together or repeated, describe a pattern worth examining before anyone signs.
Early detection, not autopsy
The sequence the sector describes is short and deliberate: SECOP I and SECOP II feed a data cross-reference, the cross-reference produces red flags, and red flags trigger preventive action. The last word is the one that matters. A system that merely documents what happened produces reports; a system that intervenes while the process is still open produces different contracts. That is the distinction between auditing and preventing, and it is the one the Ministry chose when it placed Helios inside a framework summarized in three verbs: prevent, detect, act.
Why the timing matters
Helios did not arrive at an arbitrary moment. The Defense Sector is executing the largest capability acquisition plan in its history, with funding that extends through 2032. In that setting the risk is not theoretical: every procurement line is a target for private interests. An early-detection system does not eliminate that risk, but it changes the calculation for anyone attempting it, because it lowers the odds that an irregularity will go unnoticed long enough to take hold.


