Head of AI

OSavul, Inc.; United States; Remote; USD 180,000 - 280,000; Full time

Benefits offered by this job

Fully remote work

Flexible time-off

Mission-driven startup

Job summary

OSavul, Inc. seeks a Head of AI to lead the AI function behind our intelligence platform, owning its direction across every AI track.

This player-coach role will lead and grow a team of AI engineers, while staying hands-on with architecture, code, and deployments for restricted environments. You will collaborate with the CEO, product, and data engineering to shape AI strategy, security, and execution, and to deliver architecture that scales with client needs.

Qualifications

• Proven experience leading AI/ML teams in a player-coach capacity, while remaining hands-on in architecture and code.
• Track record of owning an AI roadmap and taking AI systems from concept to production.
• Ability to justify architectural decisions and explain them to technical and non-technical audiences.
• 6+ years of hands-on experience in AI/ML, with significant work on LLM-based production systems.
• Experience building agentic platforms including orchestration, memory, and evaluation.

Responsibilities

• Own the AI roadmap and architecture across all AI tracks.
• Lead the build of the agentic platform, including orchestration, tool use, retrieval, and evaluation.
• Set engineering standards for production AI: evaluation, monitoring, reliability, and cost control.
• Lead, mentor, and grow the AI team, including hiring and shaping workflows.

Skills

AI leadership

Python

LLM deployment

On-prem deployment

System architecture

Team mentoring

Stakeholder communication

English (B2)

Tools

Neo4j

TigerGraph

Neptune

vLLM

Ollama

TGI

Job description

Are you passionate about safeguarding countries, societies, and businesses from hybrid threats? Our rapidly scaling company is committed to this crucial mission, collaborating with some of the world's most influential organisations, including NATO and the EU. If you're ready to join a high-impact team shaping the future of intelligence and security, we invite you to become part of our journey. Join us and help defend the digital world with purpose and innovation!

Role Overview

We are looking for a Head of AI to lead the AI function behind our intelligence platform and own its direction across every AI track. This is a player-coach role. You'll lead and grow a team of AI engineers, and you'll stay hands-on: designing architecture, writing code, and making the hard technical calls yourself. Many of our clients work in restricted environments, so you'll bring real experience running models on-prem and locally, not only through hosted APIs. You'll work closely with the CEO, product, and data engineering, with real ownership over AI strategy, architecture, and execution.

Key Responsibilities

AI Strategy & Architecture

• Own the AI roadmap and architecture across all AI tracks, from enrichment and knowledge modelling to agentic systems in the product.
• Define how ontologies, knowledge graphs, and models fit together, and how agents reason over connected data.
• Set the model strategy: when to use hosted models, when to run open models on-prem or locally, and when to fine-tune, balancing quality, cost, and deployment constraints.
• Own the security architecture of our AI systems, from guardrails and input/output controls to sandboxed execution for agents and tools.

Hands-on Delivery

• Stay hands-on in design and code, especially on the hardest architectural problems.
• Lead the build of our agentic platform, including orchestration, tool use, retrieval, and evaluation.
• Set engineering standards for production AI: evaluation, monitoring, reliability, and cost control.

Team & Function Building

• Lead, mentor, and grow the AI team, including hiring and shaping how the team works.
• Translate business and client needs into a clear AI plan, and communicate trade-offs to leadership and non-technical stakeholders.

Requirements

Leadership

• Proven experience leading AI/ML teams in a player-coach capacity, while remaining hands-on in architecture and code.
• Track record of owning an AI roadmap and taking AI systems from concept to production.
• Ability to make and justify architectural and technology decisions, and to explain them clearly to technical and non-technical audiences.
• 6+ years of hands-on experience in AI/ML, including significant work on LLM-based systems in production.
• Proven experience building agentic platforms, beyond individual agents or chatbots: orchestration, tool use, memory, and evaluation.
• Hands-on experience deploying and running on-prem or local models (e.g. vLLM, Ollama, TGI), including serving, quantisation, and performance tuning.
• Experience designing ontologies and data models for structured knowledge.
• Production experience with graph databases (e.g. Neo4j, TigerGraph, Neptune) and using graph data with LLMs.
• Proven experience building data enrichment pipelines with LLMs: extraction, classification, and entity resolution at scale.
• Strong Python skills and solid grounding in transformer architectures, fine-tuning, and evaluation methodologies.
• Hands-on experience implementing state-of-the-art security for AI systems: guardrails, prompt-injection and jailbreak defence, access controls for agent tools and data, and sandboxed execution environments.
• At least B2 English level.

We Offer

• Opportunity to contribute to a mission-driven startup and support prestigious clients, including governments worldwide, enterprise clients, and leading NGOs in tackling global security challenges.
• Autonomy and freedom to drive experiments and bring your own ideas to life.
• Flexibility of fully remote work.
• Flexible, unlimited time-off policy.

Given the large number of applications we process, we are unfortunately unable to provide individual updates to every applicant. If you don't hear back from us, please consider that your application was not selected for this role.

We truly appreciate your time and interest in applying to join our team!

Excited to work together on your next project?

Full-cycle platform for hybrid risks

Sabotage, espionage, cyber&influence operations and physical threats. We deploy our AI technologies and experts to protect people, businesses and states from emerging threats.

10+

Data Types

1billion

Data points analysed monthly

100+

Countries covered

AI-powered hybrid threat intelligence

Osavuls tracks ongoing hybrid threats worldwide to develop best-in-class knowledge about threat actors, emerging trends, vulnerabilities, tactics and procedures. Prepare your defences in advance using hybrid threat intel.

Alerting and early-warning on hybrid threats

Get alerted as soon as you are targeted with a hybrid threat. Real-time monitoring by AI and security experts.

Risk modelling to preempt risks

Know Who, How, Where and When may attack next. Let AI analyze your security posture, hostile actors’ intent and capabilities to predict the next attack.

Mitigation and disruption

Preemptive measures, mitigation plans and urgent takedown services - disrupt hostile hybrid activities with Osavuls technologies and experts.