Data Scientist – AI Systems & OCR

About the role

JiBe is seeking a Data Scientist to design and implement advanced AI systems that power its cloud‑based ERP for the shipping industry. You will work on cutting‑edge agentic AI, OCR, and Retrieval‑Augmented Generation solutions that automate complex business workflows.

Key responsibilities

• Develop agentic AI systems that orchestrate multi‑step reasoning, decision‑making, and tool use.
• Integrate large language models (via OpenRouter), fine‑tuned transformers, and classical ML models into cohesive workflows.
• Design, improve, and maintain OCR pipelines for diverse document types, including classification and information extraction.
• Contribute to Retrieval‑Augmented Generation pipelines, collaborating with data engineers on vector stores and data preparation.
• Collaborate with data engineers to ensure data from Databricks and MongoDB meets model requirements, and participate in code reviews and knowledge sharing.

Required profile

• Strong Python expertise, including async/await, decorators, context managers, metaprogramming, type hinting, and design patterns.
• Experience building production‑grade microservices with FastAPI, API design, dependency injection, middleware, async endpoints, and OpenAPI/Swagger documentation.
• Proficiency with the ML/AI ecosystem: Hugging Face Transformers, scikit‑learn, PyTorch or TensorFlow.
• Familiarity with MLflow, Docker, Kubernetes, and model serving.
• Hands‑on experience with OCR, document classification, and information extraction pipelines.

Required skills

• Python
• FastAPI
• Hugging Face Transformers
• scikit‑learn
• PyTorch or TensorFlow
• MLflow
• Docker
• Kubernetes
• Databricks
• MongoDB
• OpenAPI/Swagger
• Async/await concurrency
• OCR and document classification
• Retrieval‑Augmented Generation