AI Engineer

Quick Facts

• Role: AI Engineer
• Focus: Building “Underwriter OS” from scratch with AI agents for underwriting document automation and decision support
• Tech stack: LangGraph/LangChain, Next.js/React, FastAPI/Python, Azure, Docker/Kubernetes, CI/CD

Description

You will design and develop Underwriter OS, including orchestration of AI agents and integration with business systems used across insurance underwriting. The solution will extract and process data from broker documents using NLP/OCR, translate underwriting guidelines into accurate agent prompts, and support underwriting decision-making in a large financial institution.

Responsibilities

• Integrate Underwriter OS with external business systems (policy administration, rating engines, DMS, CRM, broker portal, Outlook)
• Build and orchestrate AI agents using LangGraph/LangChain, including coordinator-based agent execution
• Develop frontend and backend layers (Next.js/React and FastAPI/Python)
• Implement document data extraction and processing with NLP/OCR
• Deploy on Azure using Docker/Kubernetes and CI/CD pipelines
• Convert underwriting guidelines into precise instructions/prompts for AI agents

Requirements

• Minimum 4 years of experience in Data Science/AI
• At least 1 year of production deployments of agent- or LLM-based systems
• Practical Azure experience, including Azure AI Foundry, Azure Blob Storage, and Document Intelligence
• Knowledge of MLOps: MLflow, Docker, Kubernetes, and CI/CD
• Very good knowledge of LangGraph (or similar orchestration tools)
• Very good prompt engineering and context engineering
• Very good knowledge of React/Next.js and Python/FastAPI
• Fluent Polish

Benefits

• Project-based B2B cooperation with milestone-based billing and acceptance protocols
• Fast and remote application process; decision made by the project owner
• Administrative and settlement support on the Connectis side
• Participation in industry events and tech meetups
• Approx. 3 visits per week to Warsaw for onsite work
• Referral bonus: 5,000 PLN for recommendations to projects
• Flexible B2B model and quick remote workflow