AI/ML Solutions Architect

Quick Facts

• Role: AI/ML Solutions Architect (Platform Architect – AI)

• Work: Full-time; remote with in-person visits 2x/week to the Warsaw office

• Project: AI Visual Similarity Search to automatically verify product display compliance vs. planograms

Description

Design and architect end-to-end AI/ML solutions for a Visual Similarity Search use case. You will build computer vision and vector search pipelines, select models and technologies, and define API quality contracts. You’ll also oversee architecture for the AI/ML engineer, set up monitoring and model versioning/rollback, and collaborate with AI Security for GDPR and EU AI Act compliance.

Responsibilities

• Design end-to-end AI/ML architecture

• Design computer vision pipeline and vector search

• Select models/technologies; optimize costs; decide build vs buy; integrate with backend systems and RELEX

• Define API contracts and quality thresholds

• Architecturally supervise AI/ML engineering work

• Design monitoring, model versioning, and rollback procedures

• Analyze open-source licenses

• Collaborate with AI Security on security/compliance (GDPR, EU AI Act)

• Document architecture and mentor the team

Requirements

• 5+ years in software architecture, including 2–3+ years focused on AI/ML systems

• Strong ML and deep learning fundamentals; knowledge of modern NLP/LLM architectures

• Practical experience with cloud AI/ML platforms (AWS/Azure/GCP)

• MLOps experience (MLflow, Kubeflow, SageMaker) and CI/CD for ML pipelines

• Python proficiency; experience with PyTorch/TensorFlow and Hugging Face

• Experience with vector databases, RAG pipelines, and agentic AI frameworks

• Knowledge of data architecture, APIs, and integration patterns

• Ability to explain complex technical topics to non-technical people

• Experience leading technical teams or mentoring engineers

• Visual Similarity Search specifics:

  • Production-grade computer vision pipelines (object detection, embedding models, vector/hybrid search)

  • Experience with DETR/YOLO model ecosystem; Roboflow (or equivalent); Pinecone (or equivalent vector DB)

  • Ability to assess cost trade-offs for self-hosted GPU vs managed/SaaS inference

  • Knowledge of open-source license limitations (AGPL, Apache, MIT)

  • Experience with multi-signal retrieval/ranking

  • Practical AWS ML/GPU knowledge (SageMaker, EC2 GPU, Bedrock) and ability to discuss architecture with DevOps

  • Plus: petrol/automotive/retail domain experience (product recognition, planograms, store display compliance)

Benefits

• High autonomy for remote work

• Long-term stable employment

• Opportunity to work on key projects for a large, stable-sector company

• Attractive compensation aligned with skills and experience

• Benefits: Lux Med, MultiSport