AI Engineer

About the Position

DOS MIND, DOS Design S.p.a., is seeking an AI Engineer with proven experience in developing LLM applications in production. The selected candidate will be responsible for creating GenAI solutions for our clients and developing internal tooling, ensuring end-to-end ownership, from design to deployment. Additionally, by collaborating closely with the Discovery team, they will contribute to the ideation and development of new products.

Responsibilities

• Design and develop end-to-end GenAI applications (RAG, agents, function calling), and manage their production deployment
• Independently manage (ownership) features: technical design, implementation, deployment in cloud or on-premise environments, and monitoring
• Optimize quality, costs, and latency: model selection and comparison, caching, and optimization of retrieval and prompts
• Develop automatic evaluation pipelines for measuring the quality of outputs (accuracy, hallucination, groundedness)
• Integrate AI solutions with internal and client data systems
• Support the professional growth of junior colleagues through pairing and code review activities

Requirements

• Proven professional experience, estimated between 2 and 5 years, in software development or Machine Learning Engineering, with at least one year dedicated specifically to LLM applications in production contexts
• Solid experience in implementing end-to-end RAG solutions, with particular reference to chunking, embedding, hybrid retrieval, and re-ranking methodologies
• Proficiency in using at least one vector database in a production environment (such as Milvus, pgvector, Qdrant, or similar)
• Strong skills in using the Python language, the FastAPI framework (or equivalent tools), and in developing microservices-oriented architectures
• Mastery of Docker and proven ability to deploy on cloud platforms of major hyperscalers (AWS, Azure, or GCP)
• Previous experience in adopting agent frameworks (such as LangGraph or similar) or in designing function and tool calling systems
• In-depth knowledge of evaluation methodologies and observability tools dedicated to Large Language Models (including Langfuse, LangSmith, Ragas, or equivalent applications)

Nice to Have

• Proven experience in developing and managing projects related to classical machine learning (such as computer vision and predictive models) or in the field of data engineering
• Knowledge of fundamental MLOps methodologies, with particular reference to CI/CD processes applied to models and the structuring of training and serving pipelines
• Expertise in implementing self-hosted serving solutions for Large Language Models (LLM), using tools such as vLLM and Ollama, in on-premise infrastructure contexts
• Proficiency in using the Kubernetes platform for container orchestration

We Offer

• Competitive salary
• Opportunity for professional growth and development
• Collaborative and innovative work environment

About the Company

DOS is a design company of over 100 professionals, including designers, facilitators, and researchers, committed to producing a positive impact in large organizations.

Application Process

• First individual interview: Can be in-person or remote. You will meet with our People Management Process Area: we will get to know each other and evaluate your soft skills, relational, character, motivational aspects, and alignment with our values.
• Second individual technical interview: Preferably in-person. You will meet with the Head of the Process Area you will be part of: we will delve into your hard skills and take the time to discuss our projects. In this phase, you may perform technical exercises related to your area of expertise.
• Third individual in-depth interview: Exclusively in-person and is the final phase of the selection process. It is important for us because we will further explore your hard and soft skills and show you your development and training path at DOS; it is important for you because you can share your expectations regarding economic and organizational aspects.