AI Operations Engineer

OpenSpring; Lisboa; Presencial; EUR 50 000 - 70 000; Tempo integral

Resumo da oferta

OpenSpring in Lisbon is looking for a Machine Learning Operations / Language Model Operations Engineer. This role involves designing and managing infrastructure for AI systems, ensuring reliability and efficiency in data handling.

Candidates should have strong experience with production-grade data platforms and LLM observability, as well as skills in Python and data privacy controls. Join our innovative team in Lisbon and make an impact!

Qualificações

• Proven experience building and operating production-grade data platforms.
• Hands-on experience developing ML/LLM evaluation systems.
• Strong understanding of LLM observability and debugging tools.

Responsabilidades

• Design and implement scalable, source-agnostic ingestion pipelines.
• Build and maintain data processing workflows.
• Collaborate with AI, engineering, and product teams for system robustness.

Conhecimentos

Production-grade data platforms

ML/LLM evaluation systems

Data privacy controls

Production-level Python engineering

Data pipelines

Ferramentas

Azure

AWS

TypeScript

Descrição da oferta de emprego

Machine Learning Operations / Language Model Operations Engineer

As part of our continued expansion in AI-driven platforms and data-intensive systems, we are looking for a Machine Learning Operations / Language Model Operations Engineer to join our advanced engineering team. In this role, you will be responsible for designing, building, and operating the end-to-end infrastructure that supports production AI systems, ensuring reliability, observability, data quality, and continuous improvement of machine learning and LLM-based solutions.

Key Responsibilities

• Design and implement scalable, source-agnostic ingestion pipelines for production ML and LLM data.
• Build and maintain data processing workflows including ingestion, redaction, storage, classification, and slicing of production signals.
• Define and implement data storage strategies, including tiering, retention policies, and privacy-aware data handling.
• Develop observability and debugging tools, including dashboards and query systems for production AI systems.
• Implement evaluation and monitoring frameworks, including offline evaluation sets, online autoraters, and regression detection systems.
• Build automated triage systems to identify, classify, and surface production failures.
• Develop and maintain PII redaction mechanisms and enforce data governance and compliance policies at ingestion level.
• Design and operate LLM evaluation mining workflows to continuously improve model and prompt performance.
• Implement alerting systems to detect regressions across model and prompt deployments.
• Collaborate with AI, engineering, and product teams to ensure robustness and reliability of production AI systems.
• Evaluate and select tooling, infrastructure, and hosting strategies for ML/LLM platforms.
• Own the operational reliability of the entire ML/LLM data and evaluation pipeline.

Requirements

• Proven experience building and operating production‑grade data platforms or ML systems, including ingestion, storage, access control, monitoring, and on‑call responsibilities.
• Hands‑on experience developing ML/LLM evaluation systems (e.g., regression test sets, autoraters, LLM‑as‑a‑judge frameworks, or golden datasets).
• Strong understanding of LLM observability, tracing, and debugging tools.
• Experience implementing data privacy controls such as PII redaction in production environments.
• Deep understanding of failure modes in ML and LLM systems (hallucinations, retrieval failures, agent loops, ASR/TTS degradation, prompt/model regressions).
• Strong production‑level Python engineering skills, with a hands‑on mindset.
• Solid understanding of data pipelines, system reliability, and distributed data processing.

Nice-to-Have Skills

• Experience in multi‑tenant or SaaS architectures with strict data isolation requirements.
• Familiarity with Azure and/or AWS cloud ecosystems.
• Experience making infrastructure trade‑off decisions between managed services and self‑hosted solutions.
• Knowledge of vector databases, embedding techniques, and clustering or unsupervised failure detection methods.
• Experience with data versioning tools such as LakeFS, DVC, or Delta Lake.
• Familiarity with GDPR, data deletion workflows, and compliance‑driven data systems.
• Exposure to embedded, automotive, or constrained environments.
• Experience with non‑English language model evaluation or multilingual datasets.
• Experience working with LLM APIs such as Anthropic Claude, OpenAI models, or open‑source alternatives.
• Familiarity with CI/CD workflows, GitHub‑based development, and modern DevOps practices.
• Experience with dashboard development using TypeScript or similar frontend technologies.

Equal Opportunity Statement

At Tieto, we welcome applicants of all backgrounds, genders (m/f/d), and walks of life. We are committed to diversity, equity, and inclusion.