RiceF W - Internal - AI Engineer(Intelligent Agents & Model Optimization)

Overview

We’re looking for an experienced AI Engineer who can design, build, and optimize intelligent agents and machine learning systems end-to-end. This role involves architecting multi-agent systems, developing and fine-tuning models, and implementing performance optimization techniques (quantization, parallelization, pruning, etc.) across cloud environments (AWS, Azure, or GCP).

Responsibilities

• Design and develop AI-driven agents capable of autonomous reasoning, task execution, and contextual decision-making.

• Implement and fine-tune LLMs and multimodal models using techniques such as quantization, LoRA, and PEFT to optimize for latency and cost.

• Build modular inference pipelines supporting distributed training and inference (e.g., DeepSpeed, PyTorch Distributed, vLLM).

• Design feedback and reinforcement loops to improve agent reliability and task completion accuracy.

• Integrate models into production systems through scalable APIs and workflow orchestration (e.g., FastAPI, LangChain, Prefect, or Airflow).

• Leverage vector databases and RAG pipelines for contextual augmentation.

• Benchmark models and measure latency, throughput, and token-level efficiency to ensure production readiness.

• Collaborate with data engineers to ensure clean, structured, and well-governed datasets for model training and validation.

• Maintain observability using logging, telemetry, and experiment tracking (Weights & Biases, MLflow, etc.).

Required Skills

• Proficiency in Python, PyTorch, and Transformers (Hugging Face).

• Experience building and deploying custom fine-tuned or quantized LLMs.

• Hands-on with distributed computing, GPU acceleration, and model compression.

• Familiarity with LangChain, OpenAI/Anthropic APIs, or similar agent frameworks.

• Cloud deployment expertise on AWS (SageMaker, Bedrock), Azure ML, or GCP Vertex AI.

• Strong understanding of data pipelines, version control, and DevOps for ML (MLOps).d

Preferred Qualifications

• Experience with reinforcement learning (RLHF / RLAIF).

• Background in AI safety, model interpretability, or prompt engineering.

• Contributions to open-source AI frameworks or multi-agent systems.