AI Engineer

Deskripsi pekerjaan

Gathering, cleaning, and labeling large datasets so machine learning models can learn from them.

Designing, training, and fine-tuning AI models (like neural networks) from scratch.

Converting machine learning models into APIs or software so they can be seamlessly used in apps or websites.

Connecting AI models with existing back-end/front-end systems and cloud servers.

Evaluating the performance, speed, and accuracy of AI systems, and tweaking them to reduce errors.

Monitoring deployed AI systems in production to ensure they adapt to new data and changing business needs

Kualifikasi

• 3+ years of hands-on experience in developing and deploying ML models into real-world business applications or research environments.
• Strong understanding of ML/DL frameworks such as Jupyter Notebook, Anaconda, TensorFlow, Keras, Scikit-learn, PyTorch, and MXNet.
• Proven experience working with cloud service platforms (AWS, Azure, or GCP) for ML/DL pipeline orchestration including GPU-based training (CUDA), model evaluation, and deployment (e.g., SageMaker, Docker, or Vertex AI).
• Proficiency in Python and core data/ML libraries such as Pandas, NumPy, and Scikit-learn.
• Solid grasp of machine learning algorithms (classification, regression, clustering, feature selection, hyperparameter tuning, etc.).
• Experience developing and fine-tuning Large Language Models (LLMs) for text, code, or image generation.
• Understanding of Retrieval-Augmented Generation (RAG) architecture, including knowledge of vector databases (e.g., FAISS, ChromaDB, Milvus) and embedding models.
• Experience integrating LLMs with external tools, APIs, and data sources through frameworks such as LangChain, LlamaIndex, or similar orchestration layers.
• Familiarity with MCP (Model Context Protocol) or modern context-sharing protocols for building scalable, composable AI systems.
• Experience implementing modern AI pipelines involving fine-tuning, prompt engineering, context retrieval, and model evaluation workflows.
• Knowledge of model optimization and monitoring (latency, throughput, token efficiency, and hallucination detection).
• Ability to collaborate with cross-functional teams (data engineers, analysts, and software developers) to deliver AI-powered features and applications.