I build security for the age of AI agents.
I'm Gustavo Aragón, founder of Oktsec. I've spent 20+ years building technology across Latin America and global markets, from regulated fintech to AI agent security.
I came to AI agent security through fintech and regulated markets, where permissions, auditability, operational controls and trust boundaries are not optional. I have spent more than 20 years building technology across New Zealand, Hong Kong, São Paulo and Buenos Aires, including 12 years in CTO roles. In Brazil, I spent seven years building companies, including Brelo, a fintech I co-founded in São Paulo.
Talk about advisory →
Gus AragónFounder, Oktsec
Track record
20+
years building software
12
years in CTO / product leadership
250+
security findings reported
Recognition & affiliations
Claude Partner Badge: Claude Code
Oktsec is part of Anthropic's Claude Partner Network. The badge validates end-to-end readiness to scope, deploy, configure and run Claude Code client engagements, assessed on real deliverables: configuration packs, deployment decision docs, security questionnaires, custom commands, Skills, hooks and telemetry configs.
Agentic AI Foundation Ambassador 2026
June 2026, issued by The Linux Foundation. The Agentic AI Foundation is a Linux Foundation project advancing open and interoperable infrastructure for agentic AI. The ambassador role focuses on developer education, practical adoption and contribution pathways around AAIF-hosted projects, including MCP, Goose, AGENTS.md and agentgateway.
1st place overall · Anthropic × Kaszek LATAM hackathon
April 2026. I assembled the team behind AgentPay, which won first place overall at the first Anthropic × Kaszek hackathon in Latin America and received US$150K in Anthropic API credits. AgentPay applies a deterministic security pipeline to payments between AI agents.
Security-verified by OpenAI & Anthropic
Accepted into OpenAI's Trusted Access for Cyber and Anthropic's Cyber Verification Program, supporting advanced security research and responsible testing across frontier AI systems.
Now
Building
Oktsec, the security platform for AI agent work
Open source
Aguara + nanostack
Researching
AI agent & MCP security
Based in
Buenos Aires, Argentina
Founder–market fit
Regulated systems
I came to Oktsec from fintech and regulated markets, where permissions, auditability and trust boundaries are not optional.
Product execution
I co-founded Brelo in Brazil and scaled it past 500K users, then led product and engineering in a regulated financial group in Argentina.
Agent security
Today I build Oktsec and research agent risk across MCP, CLIs, supply chain and automation.
Career
2026–now
Founder · Oktsec
Building Oktsec, the security platform for AI agent work. Focused on policy, control and verified evidence for agent environments, plus open source tooling and security research around MCP, CLIs and automation.
2024–2026
Director of Product & Technology · Banza · Adcap Grupo Financiero
Led product and technology at a regulated financial group in Argentina, from CTO of Banza to product and engineering leadership across the group.
2017–2023
CTO & Co-founder · Brelo · Brazil
Started Flexipag in São Paulo, a B2C smartphone-collateralized lending app, then evolved the stack into Brelo's B2B platform for banks and fintechs, scaling past 500K users across LATAM.
2015–2016
CTO · Teatrix · Argentina
Led the team behind Argentina's leading digital theater platform, a subscription service with 400+ premium productions across the region.
Global experience
Buenos AiresArgentina
Bahía BlancaArgentina
São PauloBrazil
San FranciscoUnited States
AucklandNew Zealand
SydneyAustralia
BangkokThailand
Hong KongHong Kong
SingaporeSingapore
Principles
Security from day one
Built in from the first commit, not bolted on after the incident.
Product over process
The goal is something people can use safely, not a checklist.
Open by default
Tools the ecosystem can inspect, trust and improve.