Вакансия из Hirify Global, списка международных tech-компаний Для мэтча и отклика нужен Plus
Пожаловаться
Зарплата и рынок
ЗП не указана
На похожих ролях по рынку
$79к$104кмедиана$163к
48 в выборке
На СНГ рынке: $4.1к/мес ($3к - $4.8к)
88
Хорошая вакансия
развернуть
Это четко очерченная senior-позиция в AI-платформе зрелой медицинской продуктовой компании с современным стеком и хорошими льготами; без зарплаты сложнее оценить уровень компенсации относительно рынка.
Кликните для подробной информации
Зарплата не указанаСовременный стекЧеткая зона ответственностиAI в здравоохраненииХороший пакет льгот
Оценка от Hirify AI
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Senior Agents Platform Engineer (AI): Building a secure, compliant platform for production AI agents, including a shared runtime, GenAI gateway, and observability and evaluation infrastructure with an accent on reliability, healthcare data safety, and scalable LLM operations. Focus on designing agent registries, operating long-running streaming interactions, managing provider routing and failover, and implementing progressive rollouts, audit trails, and SLO-driven incident response.
Location: Paris, France; hybrid work with up to 2 remote days per week
Company
hirify.global builds cloud-native healthcare products and AI-powered solutions for patients and care teams across Europe.
What you will do
Design and own the declarative agent manifest and registry covering skills, tools, memory, models, guardrails, sandboxing, and security.
Own the LiteLLM-based GenAI gateway, including routing, failover, metering, cost attribution, capacity planning, and guardrails.
Build reliable infrastructure for long-running and streaming agent interactions, including queueing, retries, approvals, checkpointing, and resume capabilities.
Implement progressive rollouts, feature flags, emergency kill-switches, and audit trails for agents handling sensitive healthcare data.
Define SLOs, lead on-call and incident response, and build observability and evaluation systems with traces, logs, offline evaluations, and CI gating.
Ship reference agents and advise feature teams on agent architecture and platform trade-offs.
Requirements
5+ years of software engineering experience, including production experience with LLM-based agent systems.
Knowledge of agent design patterns, tool calling, state and memory management, multi-agent coordination, and human-in-the-loop workflows.
Experience with MCP, A2A, AG-UI, asynchronous control planes, queueing, retries, checkpointing, feature flags, and progressive rollouts.
Experience managing LLM routing, caching, cost, and reserved capacity with managed model providers.
Production reliability, security, sandboxing, permission scoping, and audit-trail experience for systems handling sensitive data.
Fluent English required.
Nice to have
Experience with long-term memory, artifact stores, RAG pipelines, or production feedback loops for evaluation datasets.
Experience with Kubernetes-native abstractions, CRDs, operators, or durable execution engines such as Temporal.
Experience in regulated sectors and with GDPR, HDS, ISO 27001, or SOC 2.
Culture & Benefits
Cloud-native technology platform supporting web and mobile healthcare products across multiple languages and countries.
Health insurance for employees and children, 25 days of paid vacation, and up to 14 RTT days.
Work from abroad for up to 10 days per year and relocation support for international mobility.
Mental health and coaching services, transport reimbursement, lunch vouchers, and sports or creative activity subsidies.
Access to AI coding tools, dedicated training, and a long-term employee value-sharing plan.
Hiring process
HR phone interview, followed by a hiring manager interview.
Case study and case restitution, followed by a behavioral interview.
At least one reference check.
Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →