Open Position · 2026

AI Infrastructure
Engineer

RemoteFull-time / ContractImmediate

We are building The Octogent — a fully autonomous AI agent that runs locally on any machine, executes up to 8 parallel tasks indefinitely, and requires zero API credits. We are looking for an infrastructure engineer who obsesses over low-latency inference, autonomous orchestration, and systems that run without human intervention.

What you will build
  • Design and maintain the local inference runtime powering the agent across consumer and server hardware
  • Build and optimize multi-agent orchestration layers supporting 8 concurrent autonomous task threads
  • Develop the tool-calling, memory, and context management systems that keep the agent coherent across long sessions
  • Architect zero-cost inference pipelines using quantized local models (GGUF, AWQ, GPTQ)
  • Implement the autonomous buyback and burn system integrated with Pump.fun Agentic Mode
  • Own reliability, observability, and self-healing mechanisms for unattended agent runs
What we are looking for
  • Deep experience running local LLM inference: llama.cpp, Ollama, vLLM, or equivalent
  • Strong Python (asyncio, multiprocessing) and systems programming background
  • Experience building autonomous agent pipelines (LangGraph, AutoGen, custom orchestration)
  • Understanding of model quantization, hardware-specific optimization (CPU/GPU/NPU)
  • Comfortable shipping production systems with minimal oversight
  • Solana / on-chain integration experience is a strong plus
Apply Now
Takes about 5 minutes. We read every application.
Step 01 — Identity
01 / 03