AI Infrastructure Architect

Location: One weekly visit to either the Kraków or Warsaw office is required during the first three months. Further arrangements to be confirmed.

Salary: 190 – 230 PLN net/h (B2B via Experis)

Responsibilities:

As an AI Infrastructure Architect, you will be responsible for architecting and building custom Artificial Intelligence infrastructure and hardware solutions. Your day-to-day activities will include:

• Optimizing AI infrastructure and hardware performance, power consumption, cost, and scalability of the computational stack.

• Advising on AI infrastructure technology and vendor evaluation, selection, and full stack integration.

• Acting as a subject matter expert, collaborating with and managing the team to ensure effective performance.

• Leading team decisions, engaging with multiple teams, and contributing to key strategic choices.

• Providing solutions to problems that apply across multiple teams.

• Designing innovative AI infrastructure solutions tailored to diverse client needs.

• Conducting thorough evaluations of emerging AI technologies and recommending best practices.

• Leading cross-functional teams in the implementation of AI infrastructure projects.

• Monitoring and analyzing the performance of AI systems to ensure optimal operation.

• Providing training and support to team members on AI infrastructure best practices.

Requirements:

• 3–6+ years of hands-on experience building ML/AI systems; minimum 2+ years with modern LLM or vector-search stacks.

• Proven experience with Azure services: Azure OpenAI, Azure Cognitive Search (or vector store on Azure), Azure Data Factory, Azure Databricks, AKS/ACA, Key Vault, App Config.

• Practical experience generating and working with vector embeddings and semantic search (index design, similarity search algorithms, ANN libraries).

• Strong programming skills in Python (preferred), familiarity with ML libraries (transformers, sentence-transformers, scikit-learn, PyTorch/TensorFlow).

• Experience with prompt engineering and building production-grade prompts, including safety/guardrails and prompt templates.

• Experience with data engineering: ETL/ELT pipelines, streaming/batch processing, data quality, and metadata management.

• Knowledge of OCR and image processing (Tesseract, Azure Form Recognizer, OpenCV, or equivalent) for extracting text and features from images and documents.

Nice-to-Have:

• Prior experience with Semantic Kernel, agent frameworks, or multi-agent orchestration.

• Familiarity with retrieval-augmented generation (RAG) implementations and hybrid search (BM25 + semantic).

Our offer:

• MultiSport Plus

• Group insurance

• Medicover Premium

• e-learning platform