Dastellar

We are looking for a Senior Data Scientist who can deliver end-to-end solutions, communicate both technical and business aspects to clients, and take ownership of production outcomes — helping Dastellar fulfill its mission to build custom AI/ML/Data Science solutions for businesses.

General Requirements:

• 5+ years in ML/Data Science.
• Strong experience with classical ML: sklearn, XGBoost/LightGBM, time-series models (ARIMA/Prophet/ETS or deep models), and recommender systems.
• Python 3.8+ proficiency; solid Pandas/NumPy; production-level experience with PyTorch or TensorFlow.
• Experience deploying ML models (Docker, CI/CD); basic cloud skills (AWS/GCP/Azure) preferred.
• Ability to communicate technical concepts to clients and translate risks into business impact.
• Autonomous and proactive; history of leading end-to-end tasks.
• English: B2+.
   

Nice-to-have / Growth Opportunities:

• LLM/GenAI experience: fine-tuning GPT/LLaMA/Hugging Face, building RAG/embedding pipelines, LangChain.
• Vector DBs: FAISS, Qdrant, Pinecone.
• MLOps basics: MLflow, W&B, CI/CD pipelines.
• Big Data: Spark, advanced SQL optimization.
• Domain expertise: energy or retail.
   

Main Responsibilities:

• Build classical ML models (classification, regression, clustering, recommendation, time-series).
• Data cleaning, validation, feature engineering, ETL/ELT pipelines.
• Design and deploy ML/GenAI solutions from POC to production.
• Integrate models with infrastructure (Docker, CI/CD; optionally Kubernetes/serverless).
• Monitor model performance and system health (metrics, data drift, alerts).
• Document solutions, create client presentations, and conduct R&D experiments.
• Maintain code quality: tests, reviews, reproducibility.
   

Soft Skills:

• Responsibility & ownership: end-to-end project delivery, meeting deadlines, post-release SLAs.
• Proactivity: proposing and implementing process or architecture improvements.
• Autonomy & problem-solving: translating ambiguous business goals into MVP roadmaps.
• Critical thinking: identifying data/assumption biases and proposing mitigation strategies.
   

What We Offer:

• Flexible part-time / contract work.
• Fast decision-making and low bureaucracy.
• Influence over architecture and stack choices.
• Remote collaboration and flexible schedule.
• Paid R&D participation (by agreement).