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)