AI Infrastructure & Engineering Lead

Quick Facts

• Role: AI Infrastructure & Engineering Lead

• Focus: Greenfield on-prem AI platform using dedicated NVIDIA GPU infrastructure

• Team: Lead a small Poland-based engineering team (4–5 people)

Description

Take full technical ownership of a new AI initiative: assess infrastructure, design a secure scalable on-prem environment, and deploy production-ready LLM solutions. Build and maintain the AI/ML infrastructure and model-serving stack, create data ingestion pipelines for unstructured documents and emails, and evolve an initial prototype into a reliable platform using real insurance data.

Responsibilities

• Assess existing infrastructure and design a secure, scalable on-prem AI environment with dedicated NVIDIA GPU servers

• Configure and manage GPU infrastructure (Linux, CUDA, networking, remote access, monitoring, security)

• Build and maintain AI/ML infrastructure and model-serving stack (Docker, Kubernetes, vLLM, NVIDIA Triton)

• Design data ingestion and processing for unstructured information (emails, PDFs, scanned documents)

• Develop production-ready LLM applications (RAG, fine-tuning, open-weight models such as Llama, Qwen, Mistral)

• Deliver an initial prototype with real business data and progress it to production deployment

• Build AI-powered solutions for underwriting, quote preparation, document processing, and broker communication

• Ensure data protection, access controls, and security practices

• Collaborate with international business leadership to set priorities and define the implementation roadmap

• Recruit, mentor, and lead a Poland-based engineering team (4–5 people)

Requirements

• 8+ years of professional software engineering experience, including 3+ years hands-on production AI/ML engineering

• Proven end-to-end AI/ML infrastructure delivery: assessment/architecture through production deployment

• Hands-on experience with dedicated NVIDIA GPU servers and on-premise AI infrastructure (beyond cloud-only)

• Hands-on expertise in LLMs, RAG, open-weight models, and model fine-tuning

• Solid knowledge of Linux, CUDA, Docker/Kubernetes, and model-serving technologies (vLLM, NVIDIA Triton)

• Experience building data pipelines for large volumes of unstructured documents and email data

• Strong understanding of data privacy, information security, and anonymisation practices

• Experience recruiting and leading engineering teams

• Ability to independently own technical decisions, solve complex infrastructure challenges, and communicate with business stakeholders

• Fluent English (C1+)

• Availability overlap with Bermuda business hours until ~7–8 PM Polish time

Benefits

• Greenfield AI initiative with dedicated GPU infrastructure and real business data

• Full technical ownership of architecture, infrastructure, and development roadmap

• Opportunity to build and lead your own engineering team in Poland

• Direct collaboration with international executive stakeholders

• Fully remote cooperation from Poland

• B2B compensation: $8,500–$10,000 USD/month (depending on experience)

• Potential international travel for project kick-off