AI Architect

Aegis AI Security; New York (NY); On-site; USD 150,000 - 240,000; Full time

Job summary

Aegis AI Security, led by a team of ex-Google engineers, is building a constellation of AI Agents to detect and stop adversarial AI threats. You will design the orchestration layer enabling diverse agents to collaborate on threat investigations while serving as the resident expert on model selection and customization.

We seek hands-on engineers with deep experience quantizing, pruning, distilling and serving models at scale, comfortable in a fast-paced, security-focused environment where threats

Qualifications

• Hands-on experience customizing LLM models (1-3 years).
• Experience building AI/ML systems (5+ years).
• Bonus: security/detection systems and working in adversarial environments (T&S, security, fraud).

Skills

LLM customization

AI/ML systems

Job description

We’re a team of ex-Google engineers who built some of the largest defensive platforms on the planet — Safe Browsing and reCAPTCHA. Now, we’re striking out on our own to tackle an even bigger challenge: stopping the new wave of adversarial AI attacks already hitting organizations today.

We're going after a $5B+ market, ripe for disruption. Traditional detection methods are too slow to keep up. Adversaries are using AI to craft customized, high-evasion attacks — and old-school rules-based systems don’t stand a chance.

Your Mission

Architect a constellation of AI Agents — fast, cheap, and high-precision — that work together to detect and stop emerging threats. Your job is to create the orchestration layer to enable a heterogenous set agents to collaborate to investigate threats. You’ll also be Aegis’ resident expert model selection, customization and strategy. You should have deep hands on experience quantizing, pruning, distilling, serving and orchestrating models at scale.

Why This Is Hard (a.k.a. Why It’s Fun)

• The models change daily. Literally. What worked yesterday will be obsolete tomorrow.
• The adversaries are smart — and getting smarter. We’re in a real-time arms race against AI-enabled attackers crafting tailored, evasive payloads.
• There’s no playbook. We’re building new detection paradigms as fast as the threats evolve. It’s high-stakes, high-impact work — and you’ll be in the arena from day one.

If you’re the kind of engineer who lives for tough problems and fast feedback loops, this is your kind of challenge.

Why We’ll Win

• The market is massive — $5B+ and growing fast — and incumbents are too slow to adapt.
• We’ve done this before. Our team built the core technology behind Safe Browsing (5B+ users) and reCAPTCHA (protecting 5M+ websites) while at Google.
• This is our third time building a category-defining security business. We know how to scale both the tech and the company.
• AI and security are in our DNA. We don’t bolt it on — we build it in from day one.
• We only hire top performers. Most of our team were top 1% engineers at Google — if you were the best at your last job, you’ll feel right at home here.
• We move fast. No politics. No hierarchy. Just brilliant people solving hard problems, together.

What We’re Looking For

We want engineers who are outgrowing their current roles — people who want to build at the very edge of AI and cybersecurity.

• You thrive in relentless environments where learning curves are steep and every day is a new challenge.
• You care about speed, execution, and complete thinking more than resumes or titles.
• You think in weeks and days, not months and quarters.

Despite 30+ years of combined AI experience on the founding team, we know today’s landscape changes too fast to rest on credentials. We prioritize adaptability and speed of learning above all else.

Our Culture

• Flat, flexible, and fast.
• You’ll own your decisions and your tech stack.
• You’ll have clear KPIs for success — but how you get there is up to you.
• Development cycles are measured in days, not weeks.

If you’re hungry to build AI that fights AI, and want to work with a team that moves at the speed of the real world, come talk to us.

Requirements

1-3 Years hands on experience customizing LLM models

5+ Years Building AI/ML systems

Bonus

Experience security/detection systems

Experience working in adversarial environments (T&S, security, fraud)