Artificial Intelligence Engineer

MPower Plus; Kraków; Hybrid; PLN 320,000 - 520,000; Full time

Job summary

MPower Plus in Poland is seeking an experienced ML/AI engineer to develop production-grade Python code and scalable data pipelines. The role involves building data processing solutions with Spark and Snowflake, managing ML models, and implementing MLOps practices across large datasets.

You will work with cutting-edge AI technologies, including LLM agents, RAG pipelines, and OCR/NLP tasks, collaborating with cross-functional teams to deliver robust AI-driven solutions.

Qualifications

• 8+ years of experience in ML/AI roles or equivalent.
• Experience with production-grade data engineering and ML pipelines.
• Proven ability to work with large datasets and optimize performance.
• Experience with ML model versioning and experiment tracking.

Responsibilities

• Develop and deploy production-grade Python code and data pipelines.
• Implement large-scale data processing using Spark/Snowflake.
• Build and maintain ML models and MLOps pipelines.
• Develop AI agents and RAG pipelines with LangChain or similar.
• Provide end-to-end data governance and security practices.

Skills

Python

SQL

Bash scripting

Git

MLflow

Tools

Spark

Snowflake

Airflow

Dagster

dbt

LangChain

LlamaIndex

Azure AI Foundry

Snowpark

Job description

Location: Poland, Remote /hybrid

Experience: 8+ years

Job Description

Required Qualifications

Core Programming & Tooling

• Strong programming proficiency in Python, producing production-quality, well-structured code.
• Proficient in SQL; experience with large-volume datasets in Spark and/or Snowflake.
• Comfortable scripting in Python and Bash for automation and tooling.
• Expertise with Git for version control, branching strategies, code review, and pull request workflows.
• Familiarity with experiment tracking and model versioning (e.g., MLflow).

Snowflake

• Data lake vs data warehouse architecture (medallion architecture: bronze/silver/gold)
• API integration patterns for ingesting from source systems
• Schema evolution and handling semi-structured data
• Advanced SQL (window functions, CTEs, recursive queries, semi-structured data with VARIANT/JSON)
• Warehouse sizing, scaling, and cost optimization
• Data sharing, Time Travel, Zero-Copy Cloning
• Snowpipe for streaming/continuous ingestion
• Role-based access control (RBAC), security, and data governance
• Query performance tuning (clustering keys, micro-partitions, query profiling)
• Streams and Tasks for CDC and orchestration
• Snowpark (Python/Java integration) is increasingly valuable
• Jinja templating, macros, and incremental models
• Snapshots for SCD Type 2
• Version control: Git/GitHub or Azure Repos, branching strategies
• Orchestration: Airflow, Azure Data Factory, or Dagster
• Data modeling: Dimensional modeling (Kimball), star/snowflake schemas
• CI/CD: Automated testing and deployment for dbt models and pipelines

Machine Learning & Statistical Modeling

• Hands-on experience building and deploying ML on large-scale production datasets.
• Regression, classification, clustering, and ensemble methods.
• Time series forecasting.
• Feature engineering, data wrangling, backtesting, and statistical analysis on structured data.
• Sound evaluation practice: proper train/validation/test splits, leakage avoidance, and forward/out-of-sample validation.

LLM Applications & AI Agents

• AI agent development: building agentic frameworks using Azure AI Foundry and Azure model endpoints, including tool-use patterns and multi-step orchestration.
• Experience with orchestration frameworks such as LangChain or LlamaIndex.
• Prompt engineering and chain-of-thought design for complex, multi-step workflows.
• RAG pipeline development: chunking, vector embedding/indexing (Azure AI Search), and response generation using LLM endpoints.
• Integrating LLMs and agents with structured data sources (SQL/Snowflake), tools, and external systems/APIs.
• Grounding, citation, and evaluation techniques to control hallucination and verify model outputs.

Document Processing, OCR & NLP

• OCR and document parsing using Azure Document Intelligence / Form Recognizer (or equivalent).
• Extracting and structuring information from unstructured and semi-structured documents at scale (PDFs, scans, forms, tables).
• Text processing and NLP fundamentals: entity extraction, classification, summarization, and semantic search.
• LLM-based extraction, comparison, and term/clause analysis across long documents.
• Handling document layout, multi-page context, and noisy/low-quality inputs.
• Docker: building, optimizing, and managing container images for ML and AI applications.
• Azure Container Apps (ACA): containerized deployment, KEDA-based autoscaling, and replica management.
• Azure Container Registry (ACR): image management, automated builds, and vulnerability scanning.
• Familiarity with serverless compute such as Azure Functions for targeted tasks.

MLOps, CI/CD & Deployment

• Hands-on DevOps/MLOps experience in production environments.
• CI/CD pipeline design and maintenance for ML and LLM workloads using GitHub Actions or Azure DevOps.
• Infrastructure as Code with Terraform or Bicep for reproducible Azure environments.
• Managing environment promotion and configuration across dev, staging, and prod.
• Azure Machine Learning model lifecycle: model registry, automated retraining pipelines, model versioning, and endpoint management.
• Deployment strategies (blue-green, canary), rollback procedures, and automated testing for ML pipelines (integration, load, security).
• Deploying and scaling LLM-based applications and agent infrastructure on Azure, with attention to latency, throughput, and cost.

Monitoring, Observability & Delivery

• Delivering production-ready models and services with proper testing, validation, and monitoring.
• Azure Monitor and Application Insights: logging, tracing, custom metrics, and alerting.
• Model performance monitoring: data drift detection, prediction quality tracking, and automated alerts.
• Basic UI skills (Streamlit or Gradio) for quick prototyping and stakeholder demos.

• Setting engineering standards (code review bar, testing standards) across both DE and DS work
• Ability to review code/PRs and mentor junior engineers
• Stakeholder communication (translating technical concepts to business users)
• Documentation and knowledge-sharing practices
• Cost optimization mindset (especially critical in Snowflake/Azure environments)
• Cross-functional collaboration with analytics, BI, and data science teams
• Incident management and troubleshooting production data issues