AI / LLM Ops Engineer

Agentica; United States; Hybrid; USD 57,000 - 80,000; Full time

Benefits offered by this job

Fully remote option

Hybrid work from Warsaw office

Job summary

Agentica is an AI automation startup that designs and deploys custom AI products and agents for B2B clients worldwide. You’ll be the hands that build our AI delivery: you take concrete AI projects—including agents, RAG pipelines, chatbot engines, and autonomous automations—and drive them from scoping through delivery, while our tech team oversees architecture.

This is production work where quality, reliability, and data sensitivity matter.

Qualifications

• Experience building AI agents with LangChain / LangGraph.
• Practical experience with RAG systems.
• Experience building chatbot engines (session durability, persistence).
• Familiarity with autonomous/unattended workflows.
• Knowledge of open-source/open-weight models and on-premise deployment.

Responsibilities

• Own client AI projects end to end from scoping through delivery.
• Build and maintain AI agents and orchestration flows with Python, LangChain / LangGraph.
• Design and operate chatbot engines with session handling and state management.
• Develop autonomous automations for scheduled reports and overnight processing.
• Keep LLM-powered services healthy in production: latency, costs, quality.

Skills

Python

LangChain

LangGraph

RAG systems

Chatbot engines

Autonomous automation

Open-source models

Git

Docker

CI/CD

Tools

Git

Docker

GitHub Actions

Job description

About Agentica

Agentica is an AI automation startup that designs and deploys custom AI products and agents for B2B clients worldwide. The company builds production-ready AI systems that automate and replace manual business processes across helpdesks, internal tools, e-commerce, production, and other high-friction areas. Its team delivers practical agentic AI solutions for clients ranging from cybersecurity software distributors to global Web3 companies.

The Role

You’ll be the hands that build our AI delivery: you take concrete AI projects — agents, RAG pipelines, chatbot engines, autonomous automations — and drive them from scoping through delivery, while our tech team oversees architecture. This is production work where quality, reliability, and data sensitivity matter. You ship clean, maintainable code that’s ready for client review, and when a problem needs architectural depth, you escape it — that’s how the role is designed to work. Equally important: as much of the work is the background and batch side — scheduled jobs, report generation, data pipelines that run unattended — as the chatbots and agents that face the client.

What You’ll Do

• Own client AI projects end to end, from scoping through delivery.
• Build and maintain AI agents and orchestration flows with Python, LangChain / LangGraph.
• Design and operate chatbot engines — session handling, persistence, state management across turns.
• Build and improve RAG systems: ingestion, chunking, embeddings, retrieval, grounding.
• Develop autonomous automations that run without a human in the loop (e.g. scheduled report generation, overnight data processing).
• Keep LLM-powered services healthy in production: monitor latency, costs, and output quality over time.
• Work with open-source / open-weight models and adapt solutions for on-premise deployments.
• Integrate AI features into the broader product — APIs, data models, and the surrounding infrastructure.
• Participate in code review and contribute to technical documentation.
• Maintain CI/CD pipelines in GitHub Actions or similar tools.

What We’re Looking For

• Python as your primary language — you reach for it by default.
• Experience building AI agents with LangChain / LangGraph.
• Practical experience with RAG systems.
• Experience building chatbot engines (session durability, persistence, state).
• Familiarity with systems that operate autonomously — unattended / scheduled AI workflows.
• Working knowledge of open-source, open-weight models and what on-premise deployment entails.
• Clear communication and a strong sense of ownership.
• An AI-first mindset and willingness to learn fast.
• Write code that a colleague can easily read and extend.
• Comfortable with Git, Docker, and CI/CD.

Nice to Have

• MLOps experience (serving, monitoring, evaluation, deployment of ML models) — a plus, not a gate.
• Experience designing evaluations for LLM outputs (test sets, prompt/agent regression testing).
• B2B SaaS or enterprise product background.
• GDPR / PII handling experience (data classification, audit trails).

What We Offer

• Attractive salary PLN 20,000–28,000 depending on experience.
• Fully remote or hybrid from our office in central Warsaw.
• Flexible working hours built around output, not presence.
• Benefits: Multisport & MedicoverHigh-ownership projects with real technical depth and steep learning curves.
• Early-stage leverage: join while the company is scaling fast and shape how it grows.