Engineering

AI Research Engineer — Agentic Systems

$2,800 – $4,800 / moFull-timeRemote (GMT+3 to GMT+10)Senior

Design the core reasoning loops of Rine Core — planning, tool use, memory, and self-evaluation — that let our agents operate autonomously for days, not minutes.

What You Will Do

Architect multi-step agent runtimes: planners, executors, critics, and reflection loops.
Implement tool-use protocols, sandboxed execution, and safe failure recovery.
Design memory systems (episodic, semantic, vector) for long-horizon tasks.
Build evaluation harnesses measuring task completion, cost, latency, and safety.
Run experiments across model providers; publish internal findings and benchmarks.
Collaborate with ML Infrastructure to productionize research prototypes.

Requirements

4+ years building production software, 2+ years with LLMs or agent frameworks.
Strong Python; working knowledge of TypeScript is a plus.
Hands-on experience with at least one of: LangGraph, OpenAI Agents SDK, CrewAI, AutoGen, or custom runtimes.
Solid grasp of prompt engineering, function calling, structured output, and eval design.
Published projects, papers, or open-source agent work is a strong signal.

Skills & Tools

PythonLLM OrchestrationAgent ArchitecturePrompt EngineeringEvals & BenchmarkingVector DBsTypeScript

What We Offer

Remote-FirstWork from anywhere in your announced time zone band. Async by default.

Meaningful EquityEvery full-time role ships with an equity package. We win together.

Learning BudgetAnnual budget for courses, conferences, compute credits, and books.

Mission TimeDedicated hours each month to point agents at a problem you personally care about.

How to Apply

Send your CV, portfolio, and a short note on why this mission is yours to hrd@rinejourney.xyz with the subject line “Application — AI Research Engineer — Agentic Systems”. Applications are reviewed on a rolling basis — no portals, no ghosting.

One click opens Gmail with a pre-filled application to our HR team.

Apply via Gmail ↗or use your default mail app