3 days ago
London, United KingdomSenior
Responsibilities
- Design, build, and run production AI agents that automate multi-step underwriting decisions over unstructured data.
- Build the agent runtime, including orchestration, durable state, retries, and recoverable long-running workflows.
- Develop tools, APIs, and data services with clear contracts, predictable failure modes, and auditability.
- Engineer guardrails, human-in-the-loop checkpoints, and fallbacks for high-stakes automated decisions.
- Instrument, monitor, and troubleshoot agents and services, including tracing, cost, latency, and decision logs.
- Drive system design and architectural decisions for agents and supporting services.
- Design and evolve data models and database schemas for submissions, agent runs, decisions, and audit trails.
- Provision and manage cloud infrastructure with Terraform.
- Collaborate with business, product, and engineering teams on delivery and engineering quality decisions.
- Support other engineers through code reviews, guidance, and knowledge sharing.
Requirements
- 5+ years of professional software development experience in Python or another backend language such as Go, Java, or C#.
- Hands-on experience building LLM-based agents or agentic systems in production, ideally with LangGraph, LangChain, or equivalent systems.
- Strong understanding of agent failure modes, including non-determinism, tool failures, runaway loops, cost, and latency issues.
- Strong experience with asynchronous and concurrent programming and unit and integration testing.
- Day-to-day experience delivering software with AI coding agents while maintaining code quality.
- Experience designing systems end to end, including service boundaries, data flows, failure modes, and architectural trade-offs.
- Strong relational database and data modeling experience, including PostgreSQL schema design, query performance, transactions, concurrency, and safe schema migrations.
- Experience building distributed asynchronous microservices using message brokers such as Azure Service Bus, including at-least-once delivery, idempotency, and dead-letter handling.
- Experience building documented APIs using FastAPI.
- Strong cloud and infrastructure-as-code experience with Azure and Terraform.
- Experience with Docker and production container platforms such as Azure Container Apps, Kubernetes, GKE, or ECS, plus YAML pipeline definitions.
- Clear communication, pragmatic judgment, independent decision-making, technical influence, and mentoring ability.
- Preferred experience includes high-stakes or regulated decision-making, collaboration with data scientists on non-deterministic systems, Databricks, or insurance.
Tech Stack
Categories
About CFC
CFC is a London‑headquartered specialist insurer that builds technology‑driven products for businesses, with particular strength in cyber and other emerging risks. Founded in 1999, it underwrites lines including cyber, intellectual property, management liability, medical malpractice, product recall, and property & casualty, distributed primarily through brokers worldwide. CFC serves customers in over 90 countries and offers coordinated incident response to help clients manage cyber incidents from notification to resolution.