Agentic Platform Engineer (AI, Python)

Userpilot, Inc; Dublin; Hybrid; EUR 90,000 - 150,000; Full time

Job summary

Userpilot, Inc. in Ireland is seeking an experienced engineer to own and evolve the AI agent platform that powers Lia, Userpilot's agent framework. You will build a Python-based service, integrate multi-turn grounding, and drive reliability, observability, and cost accounting across a platform used by external AI agents.

You will shape architecture, balance deterministic logic with AI components, and partner with teams to raise the bar on production readiness and customer impact.

Qualifications

• 3+ years building and shipping production software with ownership of systems
• Strong Python and CS fundamentals, including databases, queues, or real-time systems
• Experience with agent-like architectures, LLMs, and multi-turn workflows would be a plus

Responsibilities

• Design, build, and operate the agent platform end to end, from API surface through runtime, tools, retrieval, persistence, and observability
• Build LLM/agent features that ground reliably in customer data with proper evals, retries, and degradation
• Select tools and methods for signals (retrieval, deterministic logic, structured outputs, statistics, or LLM) and combine them effectively
• Treat evals, cost-per-call, and latency as first-class, ensuring scalable economics
• Write specs and PRDs guiding both human and AI implementation
• Contribute to agentic infrastructure documentation and architectural rules
• Review code for architectural consistency and reliability, ensuring boundaries and contracts are respected
• Raise the bar by shaping patterns, specs, and evals for others to build on

Skills

Python

CS fundamentals

Databases

Queues

Real-time systems

FastAPI

Pydantic

Async IO

Tools

Microsoft Agent Framework

LangGraph

OpenAI tooling

Job description

Userpilot, Inc. in Ireland is seeking an experienced engineer to own and evolve the AI agent platform that powers Lia, Userpilot's agent framework. You will build a Python-based service, integrate multi-turn grounding, and drive reliability, observability, and cost accounting across a platform used by external AI agents.

You will shape architecture, balance deterministic logic with AI components, and partner with teams to raise the bar on production readiness and customer impact.