Location: Bratislava, Slovakia, On-site
Type: Fulltime
Starting date: January 2026
Industry: Fintech
Compensation: 3 800 Eur (The final compensation and benefits vary according to candidate’s experience and type of contract)
Role summary
End-to-end ownership of data science projects—from problem framing and data work to modeling, deployment, and monitoring. You balance classical ML and modern LLM/RAG techniques with strong product sense.
What you’ll do:
● Problem framing: Translate business objectives into testable ML problems with measurable ROI.
● Data: Explore, clean, and engineer features from varied sources; partner with data engineering on pipelines.
● Modeling: Build and evaluate models (treebased, linear, time series, recommendation, NLP/CV as needed).
● LLMs & retrieval: Prototype and productionize LLMpowered features (prompting, finetuning, RAG, evaluation).
● MLOps: Package, deploy, and monitor models (drift, bias, performance); maintain feature stores and model registry.
● Experimentation: Design A/B tests; define metrics and guardrails.
● Documentation & enablement: Write clear docs; enable engineering/product to use models safely.
● Compliance & ethics: Apply privacy, safety, and fairness best practices.
What you’ll bring:
● 3–7+ years in applied ML or DS; strong Python (pandas, NumPy, scikitlearn), SQL.
● Experience with at least one deep learning stack (PyTorch or TensorFlow) and one orchestration tool (Airflow/Prefect).
● Familiarity with vector databases and LLM tooling (e.g., LangChain/LlamaIndex), and with MLflow/Kubeflow for tracking.
● Solid statistics and experimental design; ability to explain tradeoffs to nontechnical audiences.
● Bonus: streaming/realtime ML, reinforcement learning, or domain expertise in FinTech
Send us your CV to cv@mvacademy.sk
We will be happy to answer your questions and provide you with more information.
By sending your CV, you consent to the processing of your personal data (in accordance with Act No. 122/2013 Coll. on the Protection of Personal Data, as amended)

