The posting
- High Technical Impact: A unique professional challenge to shape production predictive AI models and agentic orchestration workflows for a stable, market-leading global enterprise.
- Competitive Compensation: Senior level base salary and annual target bonus.
- Benefits Package: Cafeteria allowance and private health insurance package.
- Flexibility: Hybrid working model requiring 3 days of office presence (Budapest) and offering 2 days of Home Office per week.
Required Experience
- Core Background: 6+ years in machine learning engineering, applied ML, or closely related software engineering roles with recent, demonstrated production delivery.
- Production ML at Scale: 2+ years building and operating production ML systems handling real traffic and business-critical decisions (not just notebooks or proof-of-concepts).
- Databricks ML & AWS Ecosystem: Production experience with Databricks ML ecosystem (MLflow, Model Serving, Feature Store, Unity Catalog) and supporting AWS services (SageMaker, Bedrock, S3, Lambda, Step Functions, ECS/EKS).
- Agentic AI Production Systems: Hands-on experience building agentic AI systems (multi-step orchestration, tool use, planning loops, memory management, human-in-the-loop patterns).
- Core Tech Stack & Software Engineering: Deep proficiency in Python and ML frameworks (PyTorch, TensorFlow, scikit-learn, XGBoost) with strong software engineering fundamentals (testing, version control, code review, CI/CD).
- LLM & Inference Experience: Production experience with LLM applications (RAG pipelines, vector databases like Pinecone/Weaviate/Chroma/pgvector, embeddings, prompt engineering, fine-tuning) and real-time/batch inference optimization (quantization, distillation, caching).
Preferred Qualifications
- Experience with multi-agent system architectures (specialization, inter-agent communication, shared state management, failure recovery).
- Production experience with model fine-tuning and RLHF/DPO alignment techniques.
- Hands-on GPU infrastructure management, distributed training, and compute optimization on AWS (EC2 GPU, SageMaker training jobs, Bedrock custom models).
- Familiarity with streaming ML (online learning, real-time feature computation, event-driven inference) and AI guardrail/safety systems (content filtering, output validation, cost controls).
- Contributions to open-source ML tooling or published applied ML work.
- MS or PhD in Computer Science, Machine Learning, or a quantitative field (practical production experience weighted equally).
- End-to-End ML Pipeline Engineering: Design, build, and maintain production ML pipelines end-to-end — feature engineering, model training, evaluation, deployment, serving, and monitoring on AWS and Databricks.
- Agentic AI Architecture: Architect and implement agentic AI orchestration systems using frameworks like LangChain, LangGraph, CrewAI, or custom layers, with production-grade reliability, observability, guardrails, planning loops, and memory management.
- Feature Stores & Data Pipelines: Build and optimize feature stores, training data pipelines, and feature engineering workflows serving both batch and real-time inference workloads.
- MLOps Infrastructure & Governance: Own MLOps infrastructure including CI/CD for models, automated retraining pipelines, A/B testing frameworks, model versioning, experiment tracking, and ML asset governance using MLflow, Unity Catalog, and Databricks Model Serving.
- LLM Integration Patterns: Design and implement LLM integration patterns including RAG architectures, prompt management systems, tool-use frameworks, vector databases, and memory/state management.
- Monitoring, Observability & Evaluation: Develop model monitoring for drift detection, performance degradation alerts, cost tracking, and automated remediation. Establish evaluation frameworks for predictive models (standard ML metrics, backtesting) and agentic systems (task completion, hallucination detection, tool-use accuracy, latency budgets).
- Technical Leadership & Collaboration: Drive build-vs-buy and framework selection decisions backed by prototypes and benchmarks. Mentor ML engineers through technical design reviews, code reviews, and architectural guidance.
