Production AI Engineer – LLMs, Agents & MLOps

Smart Bricks; San Francisco, Northern (CA, KY); Hybrid; USD 180,000 - 280,000; Full time

Job summary

Smart Bricks is an AI lab building autonomous reasoning systems for live, production environments. As an AI Engineer, you’ll work across the full model and agent stack, from research and experimentation to deployment and live monitoring, ensuring models stay accurate and reliable in real-time.

You will own end-to-end pipelines, implement agentic orchestration with LangChain, and collaborate with data engineers to optimize feature stores and serving latency in a fast-paced, implications-driven

Qualifications

• Hands-on production ML experience across training, deployment, and live monitoring.
• Experience building agentic systems and orchestration using LangChain or similar frameworks.
• Hands-on with LLMs, RAG pipelines, and vector search infrastructure.
• Proficiency in Python; comfortable with PyTorch or TensorFlow in production.
• Demonstrated ability to own features end-to-end and communicate across teams.

Responsibilities

• Build, deploy, and maintain production AI models for valuation, forecasting, scoring, and classification.
• Develop LangChain-based orchestration workflows and contribute to custom agent framework
• Design and improve RAG pipelines and vector search infrastructure for domain retrieval.
• Collaborate with data engineers to improve feature store quality and serving latency.
• Build and maintain model monitoring, drift detection, and retraining triggers.
• Contribute to full MLOps lifecycle: experiment tracking, versioning, deployment, rollback.

Skills

Production ML engineering

LangChain orchestration

LLMs / RAG pipelines

Python

Clear communication

Ownership

Tools

PyTorch

TensorFlow

Docker

Kubernetes

AWS

LangChain

Snowflake

Pinecone

Kafka

Spark

Job description

Smart Bricks is an AI lab building autonomous reasoning systems for live, production environments. As an AI Engineer, you’ll work across the full model and agent stack, from research and experimentation to deployment and live monitoring, ensuring models stay accurate and reliable in real-time.

You will own end-to-end pipelines, implement agentic orchestration with LangChain, and collaborate with data engineers to optimize feature stores and serving latency in a fast-paced, implications-driven