1 day ago
Cluj-Napoca, RomaniaSenior
Responsibilities
- Design, build, and operate production-grade LLM applications using RAG, semantic search, prompt orchestration, and agentic patterns.
- Develop reusable AI services and APIs for conversational experiences with intent detection, session context, structured outputs, tool calling, and streaming responses.
- Build content ingestion, embedding, indexing, retrieval, reranking, source attribution, and contextual search capabilities for website content.
- Create evaluation frameworks, datasets, regression testing, monitoring, feedback loops, and safeguards for AI quality, safety, grounding, prompt injection, and PII protection.
- Design and operate secure, observable, scalable AWS services and optimize them for reliability, latency, scalability, and cost.
- Collaborate with AEM, front-end, product, UX, content, architecture, analytics, security, and delivery teams to integrate AI into digital experiences.
Requirements
- Approximately 6+ years of professional software engineering experience, including meaningful recent experience delivering Generative AI applications into production.
- Strong Python engineering, API development, backend engineering, and practical cloud experience.
- Production experience building RAG and semantic-search systems, including embeddings, chunking, vector databases, hybrid search, metadata filtering, and retrieval optimization.
- Experience integrating LLMs through Amazon Bedrock, Azure OpenAI, OpenAI, or equivalent platforms and building conversational or agentic applications.
- Experience with automated testing, CI/CD, production observability, secure software development, and quantitative evaluation of LLM/RAG quality.
- Understanding of GenAI security and safety, including grounding, prompt injection, PII and data protection, and guardrails.
- Ability to collaborate across engineering, product, UX, content, architecture, analytics, and security disciplines and explain complex technical decisions.
- Preferred experience includes Amazon Bedrock and AWS AI services, OpenSearch, MLflow, Terraform, GitHub Actions, enterprise CMS platforms, AEM Cloud Service, content taxonomy, React, TypeScript, personalization or recommendation systems, and Adobe Analytics.
Benefits
- Professional development opportunities, interesting work, supportive leaders, and an inclusive culture.
- Yearly budget and flexible benefits package of up to 20% of annual salary.
- 30+ days off, including legal days, birthday, public holiday replacements, and benefits options.
- Performance bonus scheme and matching charity contributions.
- Core benefits include pension, life and medical insurance, meal vouchers, and travel insurance.
- Hybrid work with colleagues expected to work in the local office or onsite with clients at least three days per week; office-based teams have at least one weekly anchor day.
Tech Stack
Categories
About Marsh
Marsh is a global insurance broker and risk advisor that helps businesses, public entities, and individuals assess, transfer, and manage risk through insurance placement, captives, and advisory services. It sells brokerage, risk consulting, and claims management across industries, including cyber and property, and operates in more than 130 countries. Marsh is a business of Marsh McLennan and is headquartered in New York.