M

Metacognition

OS for lifelong-memory AI agents

Research Scientist

AI ResearchConfirmed live in the last 24 hours
Full-timeRemote in AustraliaNo salary listed

About the Job

Requirements

  • A PhD in Machine Learning, Computer Vision, NLP, or a related field is required.
  • A strong publication record at top AI venues is required.
  • The candidate must be able to independently drive research from idea to publication.
  • Deep technical expertise in at least one of reasoning, memory, multimodal learning, or large-scale model training is required.

Responsibilities

  • Define and pursue original research questions within the company's priority areas of memory, reasoning, planning, data, simulation, and multimodal AI.
  • Publish research at top AI venues, including NeurIPS, ICML, ICLR, CVPR, and ECCV.
  • Collaborate with research scientists and machine learning engineers to develop and validate ideas.
  • Engage with external academic collaborators and the broader AI research community.
  • Mentor PhD interns and junior research scientists.
  • Contribute to a rigorous and supportive research culture.
  • Set a research agenda in collaboration with the Chief Scientist and aligned with company priority areas.
  • Keep current with the latest research and drive work through to publication.

Desired Qualifications

  • Postdoctoral experience or equivalent independent research experience is preferred.
  • Experience mentoring students or junior researchers is preferred.

About the Company

Metacognition

OS for lifelong-memory AI agents

Metacognition AI develops AI systems with persistent memory, continual learning, long-term reasoning, and explicit knowledge representation to support always-on multi-agent systems that handle dynamic information, conflicting goals, and organisational preferences. Its Adelaide/AIML-linked team focuses on business tools that retain corporate knowledge, robots that learn from interaction, and agents trained on specialized tasks by spoken instruction. Its first product, Zeus, is an operating system for agents that sits beneath large language models to govern how they interact with machines and industrial systems. The company’s goal is to build a robotics-oriented operating system with lifelong memory and reasoning, enabling a network of agents to operate reliably within complex industrial environments.

Company size
1-10
Company stage
Seed
Total funding
$9.7M
Headquarters
Adelaide SA, Australia
Founded
2026

Recently Posted Jobs

Explore more opportunities at Metacognition

See more jobs