Full Time
1280 Month
40
Apr 3, 2026
We are building a groundbreaking - self evolving AI coding program- designed to rapidly elevate junior and mid level developers to elite engineering standards. This production grade system uses Agentic RAG, continual learning via reflection loops, hierarchical memory and safe self modification - all running exclusively inside NVIDIA NemoClaw - kernal level sandboxes.
You will work directly under - elite engineer Grok oversight - with every PR and self-update receiving rigorous senior + level review for clean, type safe, secure and maintainable code. This role offers - REAL long term growth on cutting edge agent infrastructure while you level up to elite standards.
Paid Test: $40 for 4 to 6 hrs. - After GitHub + video review selected candidates will complete short paid Task:
Set up a simple Nemo-Claw + OpenShell sandbox environment using the official one-command install and implement a simple researcher agent stub with a placeholder for Agentic RAG.
The stub should:
Use the NemoClaw sandbox for all execution (apply basic Landlock/seccomp-style policies - full policy template provided).
Include a basic CrewAI/LangGraph structure for a Researcher role
Log trajectory data (prompt-action-observation)
Return a simple JSON output with any "learned skill" placeholder
Full detailed instructions, sandbox YAML policy template, and success criteria will be provided after shortlisting.