LLMOps Architect

About the Company

Enterpret is at the forefront of AI-native applications, unlocking the power of customer feedback for businesses. We centralize feedback from every channel and transform it into actionable insights for teams at leading companies like Perplexity, Notion, Canva, and Figma.

Responsibilities

• Design and evolve the ML platform for training, serving, and retraining encoders and LLM models using AWS, Terraform, OpenAI, and Anthropic
• Build CI/CD pipelines tailored for ML, including model versioning, testing, and canary releases
• Deploy and manage model serving systems for real-time inference and batch pipelines
• Set up observability for model performance and data drift using Braintrust and custom alerts
• Lead incident response, root cause analysis, and postmortems for ML systems
• Track and optimize cloud usage for ML workflows to ensure cost-efficiency
• Implement governance and security across the stack, including IAM and data access
• Partner with ML and product teams to productionize GenAI models for Knowledge Graphs and Adaptive Taxonomy engines
• Evaluate and build tools for model registry, feature stores, and orchestration
• Champion MLOps best practices and mentor researchers transitioning into engineering roles

Requirements

• Minimum 6 years of experience in MLOps and ML infrastructure
• Deep expertise with AWS (SageMaker, EC2, EKS, S3, IAM)
• Proficiency with Infrastructure-as-Code (Terraform)
• Experience with container orchestration (Docker, Kubernetes)
• Strong Python skills
• Hands-on experience with CI/CD systems like GitHub Actions, ArgoCD, or Jenkins
• Proven ability to monitor production ML systems (drift, latency, uptime)
• Familiarity with model serving stacks and experimentation tools (MLflow, Langsmith)
• Proficiency with AI coding agents like Claude and Cursor

Preferred Qualifications

• Experience with Go, Bash, or Rust
• Exposure to GenAI workflows (LangChain, vector DBs, RAG)
• Experience with encoder/LLM model tuning and reinforcement learning
• Knowledge of responsible AI practices