Design, build, and operate large-scale AI agent workflows using LLMs, under strict latency, reliability, and cost constraints.
Own the full lifecycle of hybrid AI systems combining LLM-based agents and traditional Machine Learning models.
Architect and optimize durable, stateful workflow orchestration for long-running, high-impact review processes.
Define and execute robust evaluation strategies for LLM agents, including offline datasets, quality metrics, regression testing, and production monitoring.
Lead end-to-end LLMOps: provider selection, cost optimization, rate-limit management, model upgrades, and failover strategies.
Translate product requirements into technical designs, assess feasibility, and manage risks inherent to stochastic AI systems.
Collaborate closely with Product, Compliance, and global Engineering stakeholders to communicate system capabilities, limitations, and risk profiles.
Requirements
Proven experience building and operating LLM-based agents in large-scale production environments.
Strong understanding of agent architectures, prompt engineering, structured outputs, guardrails, cost, and latency optimization.
Hands-on experience with agent orchestration and observability frameworks (e.g., LangGraph, LangSmith, LangFuse or similar).
Solid background in traditional Machine Learning pipelines, model evaluation, and serving.
Strong software engineering skills in Python, including async programming, API design, testing, CI/CD, and cloud-native infrastructure (Docker/Kubernetes, GCP/AWS).
Practical experience with LLMOps, including cost control, rate limits, model lifecycle management, and multi-provider strategies.
Advanced or fluent English.
Bachelor's degree completed.
Tech Stack
AWS
Cloud
Docker
Google Cloud Platform
Kubernetes
Python
Benefits
We cover 100% of medical and dental plans for Sinchers and eligible dependents through Bradesco Saúde.
With the Caju flexible benefit card, our Sinchers can choose to use benefits on food, education, and home office assistance.
Our Sinchers can enjoy paid maternity leave for 180 days and paternity leave for 30 days. Plus, we provide daycare assistance for kids up to five years old.
We partner with Wellhub to help Sinchers access gyms and wellness options.
Our Partnership with Prudential life insurance provides coverage for all Sinchers in the events of unexpected absences, serious illness, accidents, and disabilities.
We offer annual reimbursements for certain expenses related to disabilities and/or transgender needs.
Our Sinchers can take a day off on their birthday to celebrate with their loved ones.