AI Engineer

About the Role

The role involves designing, developing, and deploying scalable generative AI systems while overseeing and maintaining Large Language Model (LLM) deployments for internal AI tools. Responsibilities include ensuring high availability and efficiency for internal teams, as well as training and deploying deep learning and LLM models to production environments.

Responsibilities

• Create and implement reliable Retrieval-Augmented Generation (RAG) pipelines that integrate LLMs with proprietary data, focusing on delivering accurate and fast responses.
• Design and manage sophisticated Apache Airflow Directed Acyclic Graphs (DAGs) to automate workflows such as data ingestion, cleaning, embedding generation, and model retraining.
• Oversee and optimize Vector Databases including Pinecone, Milvus, and Weaviate for effective storage and retrieval of high-dimensional embeddings.
• Maintain internal LLM infrastructure, ensuring hosting, scaling, and version control while achieving 99% uptime; efficiently manage GPU resources and optimize inference performance for AI tools.
• Lead training, fine-tuning, and deployment of deep learning and LLM models through the entire MLOps lifecycle, from experimentation phases to serving in production.

Qualifications

• Bachelor's Degree or higher in Computer Science, Software Engineering, Data Science, AI or similar discipline.
• Minimum of 8 years of professional experience in software engineering.
• Proven track record in creating and deploying LLM applications and RAG pipelines in production settings.
• Hands-on expertise in Python programming and database management.
• Experience optimizing GPU computing resources, including tools like vLLM and TensorRT.
• Contributions to open-source AI or MLOps projects.
• Advanced knowledge of model fine-tuning techniques such as LoRA and PEFT.
• Strong foundational understanding of linear algebra and statistics relevant to machine learning.

Required Skills

• Programming: Proficiency in Python is essential; Java is a plus, with additional advantage for Go and Scala.
• AI/ML Frameworks: Skilled with PyTorch, TensorFlow, and Hugging Face libraries.
• Generative AI & LLMs: Experience with RAG, LangChain, LlamaIndex, and RAG pipeline architectures. Familiarity with PEFT/LoRA fine-tuning is beneficial.
• Data Processing: Knowledge of Apache Airflow and DAG creation; familiarity with PySpark or Ray is advantageous.
• Databases: Experience managing Vector Databases like Pinecone or Milvus, alongside SQL/NoSQL databases and semantic search embedding models.
• MLOps: Competency with Docker, Kubernetes, virtual LLMs like vLLM, cloud platforms (AWS, Azure, GCP), and CI/CD pipelines. Understanding GPU management and optimization via CUDA is a plus.