Skip to content
yrs shipping 4+ since 2022
orgs cofounded 3 BAISH · ASTN · AJ
hackathons 3 1st · 2nd · mid
projects 17 6 shipped

Luca De Leo

AI safety infrastructure engineer. Fieldbuilder, operator, cyborg.

Passport:
Argentina.
Location:
Buenos Aires → London, May 2026.
Available:
Joining 80,000 Hours. Advising, writing, and hackathons otherwise.

luca@baish.com.ar · lvca.dev · linkedin · @lvcadeleo

Skills

Engineer:
Full-stack. Ship in TypeScript, Python, Rust, Swift. Build and migrate production apps with coding agents.
Fieldbuilder:
Founded Argentina's AI safety hub. Cofounded ASTN and the Alignment Journal. Facilitated BlueDot's AGI Strategy and Technical AI Safety courses.
Operator:
Ran Nonlinear's grantmaking network — 50+ funders, 500+ applications, $800K+ disbursed. Fast, legible systems.
Cyborg:
Claude Code, Cursor, agent SDKs. Built tooling for AI agents (gdoc, claude-runner, auto-gsd).
Truth-seeker:
Rationalism, effective altruism, AI alignment.

Experience

Product Engineer, Career Services · 80,000 Hours

2026 — present

Building product and engineering on the career services team. Joining from London in May.

Cofounding Director · BAISH (Buenos Aires AI Safety Hub)

Aug 2024 — 2026

Founded and directed Argentina's first AI safety hub. Handing off as I move to London.

  • 130+ members across 13 cohorts of AI Safety Fundamentals.
  • Mentored 10 participants through research projects with concrete career roadmaps.
  • Designed and built baish.com.ar.

Cofounder & Sole Engineer · Alignment Journal

Mar 2026 — present

A serious publishing venue for alignment research. Shipping the platform end-to-end.

Cofounder · ASTN

2025 — present

AI safety's career platform. One profile, smart recommendations, automatic recruiter matching.

Course Facilitator · BlueDot Impact

2025 — Mar 2026

Facilitated cohort discussions for the AGI Strategy and Technical AI Safety courses, training entrepreneurs, operators, and engineers to build AI safety interventions.

Background Researcher & Operations · AI Species

2023 — Feb 2026

AI safety documentaries. 20M+ views. Research, fact-checking, ops, content strategy.

Operations Associate · Nonlinear

May 2023 — Apr 2024
  • Built management system for the Nonlinear Network: 50+ funders, 500+ grant requests evaluated, $800K+ channeled.
  • Stood up the Nonlinear Emergency Fund for FTX-affected grantees — another $800K disbursed in three days.
  • Accounting, compliance, finance ops.

Prize & Platform Development Intern · Nonlinear

Sep 2022 — May 2023

Sole full-stack developer for the Super-Linear platform. Designed prize structures with orgs and AI experts. Automated claims system.

Research & Hackathons

  • Four Paths to Failure — 1st place, Apart Research, Jun 2025.
    Red-teamed ASI governance frameworks, identified circumvention routes, proposed policy amendments. Parlayed into a month-long governance fellowship.
  • RobustCBRN Eval — 2nd place, Apart Research, Sep 2025.
    Practical benchmark robustification toolkit for CBRN safety evals — consensus detection and statistical analysis.
  • Tabletop Agents — Apart Research, Nov 2025.
    Multi-agent framework compressing AI governance wargames from years to minutes.

Education

Computer Science (AI focus) · University of Buenos Aires

2016 — 2023

Grant · Long-Term Future Fund

2022 — 2023

One-year stipend for self-directed study in AI alignment. Contributed to alignment.dev and stampy.ai. First place at Apart Research × Epoch AI Trends hackathon.

Grant · ACX Grants (Astral Codex Ten)

2022

Funded to self-study AI safety.

Languages

English (fluent) · Spanish (native) · Portuguese (limited)

Now

Joining 80,000 Hours

Buenos Aires → London

© 2026 · Luca De Leo

lvca.dev