Head of AI

About Konvu

At Konvu, we’re on a mission to make security invisible, redefining vulnerability management for the agentic world.

As AI begins to generate, test, and deploy code autonomously, legacy security approaches break down. Konvu’s AI agents reason like security engineers: automatically triaging noise, verifying exploitability with evidence, and helping teams focus on what truly matters.

Our agents plug directly into existing scanners and workflows, no rip-and-replace, no new dashboards. We deliver evidence-backed decisions where teams already work, cutting false positives and surfacing exploitable issues with confidence.

Founded by early members of Sqreen (YC W18, acquired by Datadog), we know what it takes to turn deep technical insight into tools developers and security teams actually use.

We’ve raised $5M in Seed funding and are backed by top European and US VCs, as well as leaders from companies like Datadog, GitHub, Docker, Cloudflare, Sumo Logic, and Vanta.

What you'll do

You’ll define how advanced AI systems reason about software vulnerabilities, taking cutting-edge agentic architectures and grounding them in one of the most adversarial real-world domains: security.

You’ll collaborate with the founders and work closely with early customers. This is a role for someone who thrives on making complex systems work in production, not just in papers.

Your responsibilities will include:

• Define the Frontier of Applied AI in Security: Take the latest breakthroughs in reasoning, planning, and multi-agent systems and make them work in the wild, where reliability, interpretability, and proof matter just as much as raw capability.

• Guide Training & Evaluation: Define and drive how we deploy and evaluate reasoning models for security, choosing the right techniques to achieve high precision, reliability, and explainability.

• Architect Enterprise-Grade AI Agents: Design agents that replicate how security engineers think: reasoning about exploitability, triaging vulnerabilities, and justifying every decision with evidence.

• Invent Agentic Infrastructure: Design the orchestration, memory, and feedback loops that allow multi-agent systems to operate autonomously at scale across massive enterprise codebases.

• Bridge Research and Reality: Evaluate and productionize the newest AI and security research, often before it reaches mainstream adoption, with judgment on what’s real versus hype.

• Shape a New Discipline: Security + AI is uncharted territory. What you design here could set the reference architecture for how agentic systems secure software globally.

• Lead and Inspire: Influence a team of exceptional engineers, and set Konvu’s AI direction with autonomy, impact, and speed.

Who you are

You’ve been working on intelligent systems since before the LLM boom. You’ve built systems that think, not just systems that predict. You’re not only an exceptional researcher or engineer, you’re a builder who wants to see advanced ideas come alive in production, solving one of the hardest problems in security.

• 10+ years of software experience in applied AI, with deep hands‑on experience in applied AI, reasoning systems, or large‑scale ML infrastructure.

• Proven track record of designing, training, or deploying reasoning agents, LLMs, or advanced ML systems: ideally in domains where reliability and precision matter.

• Ability to bridge research and application: from reading a new paper or experimenting with a new model, to evaluating how it should ship in production.

• Excellent communication skills, with the ability to explain complex systems clearly to both technical and non-technical teammates.

• Curiosity about security. No prior security experience required, but a genuine drive to learn a new, adversarial domain and apply AI to it.

• Leadership mindset: experience mentoring or leading small teams, with excitement to grow and set the direction of Konvu’s AI org.

• Positive, impact‑driven attitude & motivated by building something that will protect software worldwide, not just publishing results.

• Proficient in English (the office is an English‑speaking environment).

Our hiring process

• 30‑minute video call with our CEO

• 1h video call with our CTO

• Onsite with the whole team: 3 hours total

• Reference calls

Why Work at Konvu

We’re building AI agents that redefine how security and engineering teams work. We’re assembling a world‑class team of builders who want to push what’s possible in security and AI.

Here is what it's like