Eternis is building systems for a world with highly capable AI.

We envision a world where everyone has a digital twin that understands your preferences, evolves with you, makes millions of decisions daily, and coordinates seamlessly with other twins. While power over AGI and resources consolidates into the hands of a few, we're building a different future—one where billions of digital twins govern powerful multi-agent systems with access to capital, physical resources, and real-world agency.

<aside> ❇️ What we're building:

• Digital twins for everyone
• Secure execution environments for verifiable execution, private storage, and autonomous operation
• Protocols for large-scale coordination between intelligent agents
• The most comprehensive dataset for human-aligned decision-making
• Memory architectures & open-source, self-hosted infrastructure

About Us: We're a well-funded startup ($30M raised) based in SF.

Stack: Go, Temporal, K8s, Rust, TypeScript (NextJS), AWS Nitro, Intel TDX

A more detailed description of our research efforts is available here.

</aside>

Role description

• Entrepreneurial and self-motivated research engineer with a focus on SOTA techniques in LLM fine tuning, reinforcement learning, RAG-based memory architectures for large datasets
• Implement and optimize state-of-the-art techniques from recent AI research papers in areas including:
  • Continuous learning systems across various reward regimes
  • Multi-layered memory architectures for digital twins
  • Compound AI systems
  • Multi-agent coordination and collaboration
• Rapidly prototype and validate new approaches to digital twin/personal AI capabilities
• Collaborate with research and product teams to translate research findings into practical implementations
• Run experiments, analyze results, and iterate quickly to improve system performance
• Contribute to research publications and open-source projects

Requirements

• Demonstrated track record of solving hard research problems in any technical field
• Publications in top-tier conferences or journals (any technical field)
• Exceptional ability to understand new techniques from papers and reproduce results quickly
• Strong programming skills and experience implementing complex algorithms
• High agency and self-motivation to drive projects forward independently
• Experience with machine learning frameworks and large language models
• Deep sense of urgency and commitment to solving problems at high speed
• Daily use of AI tools (Cursor, Windsurf, Continue.dev, Devin, etc.)
• Familiarity with the open source AI ecosystem (Ollama, tool calling, etc.)