ai engineer

1global; Berlin; On-site; EUR 90,000 - 130,000; Full time

Job summary

1GLOBAL ищет опытного инженера по разработке для создания центральной AI-платформы и внедрения управляемых процессов для внутренних инструментов и MCP-серверов. Вы будете отвечать за архитектуру, безопасность и масштабирование решений в быстро растущей телекоммуникационной среде.

Требуется глубокий опыт backend, интеграция LLM и владение инструментами типа LiteLLM / Portkey. Ваша роль включит настройку CI/CD и обеспечение соответствия требованиям к конфиденциальности данных.

Qualifications

• 5+ лет опыта разработки ПО, включая внедрение LLM в реальные системы.

Responsibilities

• Разработка центральной AI-платформы и дженерированная архитектура управления затратами, доступов и мониторинга.
• Проектирование архитектуры внутренних AI-инструментов и MCP-серверов, включая RAG-воронки и векторный поиск.
• Создание безопасных конвейеров инфо/картотеки для источников знаний с доступом по ролям.
• Разработка бекенд-сервисов и API для внутренних пользователей и команд продуктов.
• Определение инженерных стандартов тестирования, развёртывания, мониторинга и оценки моделей.
• Полный цикл разработки от требований до вывода в продакшн.

Skills

Python

Backend инженерия

LLM интеграция

CI/CD

Безопасность данных

Tools

LiteLLM

Portkey

AWS Bedrock

MCP сервера

Job description

Описание:

1GLOBAL is a global mobile communications provider that helps enterprises and consumer brands deliver connectivity worldwide. Its telecommunications platform, proprietary eSIM technology, and mobile core network support operations in 43 countries.

Задачи:

Design and build a central AI platform and governance capabilities, including an LLM gateway, token budgets, rate limits, key management, guardrails, audit logging, observability, and cost controls Engineer the architecture of internal AI tools and MCP servers, including RAG pipelines, vector search, LLM integration, and conversational interfaces Build secure ingestion and indexing pipelines for internal knowledge sources with appropriate access controls Develop backend services and APIs that expose AI capabilities to internal users and product teams building customer-facing features Establish engineering standards for testing, deployment, monitoring, and model evaluation Own the development lifecycle from requirements through production monitoring, shipping and iterating

Требования:

5+ Years of software engineering experience, including recent work integrating LLMs into real systems Strong backend engineering skills and proficiency in Python; able to build and ship full-stack internal tools independently Hands-on experience with AI gateways or model routing layers such as LiteLLM, Portkey, or AWS Bedrock, including token limits, guardrails, observability, and cost management Practical experience building MCP servers or strong fluency with similar tool-calling and agent integration patterns Experience using AI coding agents such as Claude Code or Codex, and defining skills, workflows, and standards for other engineers Working knowledge of RAG, vector search, and LLM evaluation Comfortable with CI/CD, containerization, Docker, Kubernetes, and infrastructure-as-code Familiarity with data security practices, including PII handling, access controls, and data classification Strong communication skills and ability to explain technical trade-offs clearly Fluent in English Self-directed and comfortable with ambiguity Будет плюсом: MCP-specific experience, deeper expertise in RAG and vector search, experience in regulated environments such as telecommunications, financial services, or ISO 27001

Условия:

Growth opportunities at a telecommunications company expanding over 50% year-on-year Exposure to major transactions affecting the future of the telecom industry Collaboration with the Board, Founders, Senior Management Team, and external advisors Professional development alongside industry experts Opportunities to work in different 1GLOBAL offices worldwide Collaborative, fast-paced, international work environment with open communication and a results-driven culture