Arvato Systems Malaysia Hiring! Full Time AI Operations Engineer in - Ricebowl

Requirements:

•  Degree in computer science, business information systems, data science, or a comparable field, or alternatively relevant professional experience in software development, system integration, DevOps, or AI engineering.
•  Experience developing and operating production-grade software or automation solutions.
•  Proficiency in at least one programming language, preferably Python or JavaScript/TypeScript.
•  Experience with REST APIs, webhooks, JSON, authentication mechanisms, and system integrations.
•  Practical experience with workflow automation platforms, preferably n8n.
•  Hands-on experience with large language models and generative AI, including integrating and operating models such as Gemma, Qwen, or comparable models.
•  Knowledge of model gateways or proxy solutions such as LiteLLM.
•  Basic knowledge of container technologies such as Docker and modern deployment and CI/CD practices.
•  Experience with logging, monitoring, and observability solutions.
•  Understanding of data protection, information security, and secure software development.
•  Good written and spoken English skills. German is an advantage but not required for collaboration within the international team.
•  Preferred (advantageous) qualifications: experience with Kubernetes or comparable orchestration platforms; knowledge of cloud environments and on-premises or hybrid operating models; experience with vector databases, embeddings, and retrieval-augmented generation; knowledge of prompt engineering, LLM evaluation, and automated quality testing; experience with agent frameworks and tool-calling architectures; knowledge of infrastructure as code; experience with databases such as PostgreSQL, MongoDB, or Redis; understanding of MLOps, LLMOps, or AIOps concepts; and experience with open-source models, model hosting, and GPU infrastructure.
•  Personal competencies: strong analytical and solution-oriented mindset; enthusiasm for testing new technologies and applying them in production environments; high level of ownership, quality awareness, and accountability; ability to explain complex technical topics in a clear and accessible way; independent, structured, and pragmatic working style; strong communication and collaboration skills; willingness to take responsibility for the production operation of developed solutions; and interest in business processes and cross-functional collaboration.

Responsibilities:

The AI Operations Engineer develops, integrates, and operates AI-based products and automation solutions for internal business processes and customer-facing workflows. The objective of the role is to automate recurring activities, support employees through intelligent assistance systems, and deliver new digital products based on large language models and modern automation platforms. The AI Operations Engineer is responsible for the entire solution lifecycle, from requirements analysis and solution design to technical implementation and reliable, secure, and scalable operations.

1. Analyze internal and customer-facing processes to identify opportunities for automation and AI enablement.
2. Design and develop AI-powered workflows, agents, assistance systems, and backend services.
3. Implement workflow automation using platforms such as n8n, and integrate large language models, APIs, databases, and existing enterprise systems.
4. Develop proofs of concept and transition successful solutions into production-ready applications, including reusable components, templates, and standardized integrations.
5. Integrate and operate open-source and commercial language models (e.g., Gemma, Qwen), managing centralized model access, routing, and governance using solutions such as LiteLLM.
6. Select suitable models based on quality, cost, latency, privacy, and use-case requirements, and develop and optimize prompts, system instructions, and structured model outputs.
7. Implement retrieval-augmented generation (RAG), tool calling, and agent-based workflows, and conduct model comparisons, evaluations, and quality tests.
8. Ensure stable, scalable, and transparent operation of AI solutions by establishing monitoring, logging, alerting, and error-handling mechanisms.
9. Monitor model usage, response quality, latency, cost, and availability; analyze and resolve technical incidents; and manage updates, releases, and continuous improvements.
10. Document architectures, interfaces, workflows, and operational procedures.
11. Implement data protection, information security, and compliance requirements, including role-based access and authorization concepts.
12. Ensure secure handling of sensitive company and customer data.
13. Address AI-specific risks such as prompt injection, data leakage, hallucinations, and misuse, and support the development of standards and governance policies for generative AI.
14. Work closely with business units, product management, IT, data protection, and information security teams to translate business requirements into technically feasible AI solutions.
15. Advise internal stakeholders; conduct workshops, technical alignment sessions, and product demonstrations; and support users during the introduction and adoption of new AI products.

Typical deliverables of this role include: production-ready AI assistants for employees or customers; automated processing and classification of requests; AI-powered document and information processing solutions; automated research, summarization, and reporting workflows; intelligent routing and decision-support systems; reusable n8n workflows and AI components; centralized and controlled model access through LiteLLM; and monitoring and governance solutions for enterprise-wide AI usage.

Success in this role is measured by: the number of AI and automation solutions deployed to production; reduction in manual processing time; degree of automation across selected processes; availability and stability of operated solutions; quality and accuracy of AI-generated outputs; adoption and usage by internal or external users; cost per automated transaction or model request; and time from initial idea to production deployment.