Generative AI engineer for multi-agent systems

Digitl; United States; Remote; USD 101,000 - 169,000; Full time

Benefits offered by this job

Co-working spaces in Europe

Remote first

Home office equipment

Job summary

Digitl ищет инженера Generative AI, который будет заниматься разработкой и внедрением многоагентных систем в продакшн. Основной фокус на архитектуры межагентного взаимодействия, обработку вывода и эксплуатацию моделей.

Требуется опыт работы с LLM в проде, Python и Google Cloud (Vertex AI). Готовность к удаленной работе в Европе, обмен знаниями и наставничество приветствуются.

Qualifications

• : Требуется 3+ года опыта в инженерии ПО/ML/бекенд системах.
• : Опыт вывода LLM в продакшн.
• : Опыт работа с SDK Google Agent Development Kit, LangChain, LangGraph.

Responsibilities

• Design and implementation of production-ready multi-agent systems with clearly defined strategies for planning, routing, memory and tool integration
• Design of robust communication architectures between agents and development of modular components for extensible multi-agent systems
• Implementation of structured outputs, schema validation and function calling
• Definition of guardrails, fallback strategies and mechanisms to mitigate hallucinations
• Design of evaluation pipelines for quality, latency and costs
• Work with Gemini, Claude and open models via APIs
• Implementation of advanced prompt engineering and system instruction design
• Optimization of model usage in terms of costs, latency and reliability
• Building and maintaining RAG pipelines using vector databases
• Implementation of observability, tracing and monitoring for LLM-based systems
• Architectural design of cloud-native AI systems on Google Cloud
• Use of Vertex AI, Agent Engine, Model Garden and related services
• Deployment of services using Vertex AI Agent Engine, Cloud Run, GKE or serverless architectures
• Implementation of CI/CD pipelines and infrastructure as code
• Ensuring a secure architecture with IAM, service accounts and data governance
• Close collaboration with the Cloud, Ad Technology and Digital Analytics teams
• Creation of technical documentation as well as internal training and knowledge materials
• Leading technical workshops and mentoring engineers in agent system design

Skills

Python

LLMOps

Agent frameworks

Google Cloud

Vertex AI

RAG

Vector databases

CI/CD

IAM

English

Tools

LangChain

LangGraph

Goog. Agent Development Kit

Job description

Описание:

Digitl is a young and innovative company in marketing and cloud technology. In a collaborative and cross-functional environment, we develop new concepts to guide our customers from various industries into a digitally shaped future. Become part of it and contribute your own fresh and forward-looking ideas as a Generative AI Engineer to help shape Digitl's further development together.

Задачи:

• Design and implementation of production-ready multi-agent systems with clearly defined strategies for planning, routing, memory and tool integration
• Design of robust communication architectures between agents and development of modular components for extensible multi-agent systems
• Implementation of structured outputs, schema validation and function calling
• Definition of guardrails, fallback strategies and mechanisms to mitigate hallucinations
• Design of evaluation pipelines for quality, latency and costs
• Work with Gemini, Claude and open models via APIs
• Implementation of advanced prompt engineering and system instruction design
• Optimization of model usage in terms of costs, latency and reliability
• Building and maintaining RAG pipelines using vector databases
• Implementation of observability, tracing and monitoring for LLM-based systems
• Architectural design of cloud-native AI systems on Google Cloud
• Use of Vertex AI, Agent Engine, Model Garden and related services
• Deployment of services using Vertex AI Agent Engine, Cloud Run, GKE or serverless architectures
• Implementation of CI/CD pipelines and infrastructure as code
• Ensuring a secure architecture with IAM, service accounts and data governance
• Close collaboration with the Cloud, Ad Technology and Digital Analytics teams
• Creation of technical documentation as well as internal training and knowledge materials
• Leading technical workshops and mentoring engineers in agent system design

Требования:

• More than 3 years of experience in software engineering, ML engineering or backend systems
• Proven experience building and deploying LLM-based systems in production
• Hands-on experience with the Google Agent Development Kit, LangChain, LangGraph or comparable agent frameworks
• Strong Python skills and sound software architecture practices
• In-depth understanding of LLMOps, including versioning, evaluation, monitoring and cost control
• Experience with RAG architectures and vector databases
• Comprehensive knowledge of Google Cloud, especially Vertex AI and cloud-native deployment patterns
• Experience designing secure and scalable distributed systems
• Hands-on experience with Google Cloud authentication, including OAuth 2.0, service accounts and IAM
• Excellent communication skills in English and German

Условия:

• Co-working – You have access to a spot in the best co-working spaces in Düsseldorf, Zurich, Barcelona or Hamburg
• Remote first – All our processes are designed for remote work. You decide whether you prefer to work from home or spend your day in one of our offices
• Home office equipment – We provide you with a standard setup for your remote working environment
• Animal lovers – If you have a dog, you can bring it to any of our offices
• Skill building – As a highly specialized technology partner, we support your development with certifications, conference attendance and training