Open position · Human-AI Operations
You encode what humans know into something machines can learn. You are the bridge between deep domain expertise and the agent workforce that runs on it.
The role
Agents don't learn from thin air — they learn from you. As Skill Architect, you take real-world expertise from a domain you know deeply and transform it into structured, teachable skills that our agent workforce can acquire, apply, and build on. You work at the intersection of knowledge engineering, prompt craft, and curriculum design. You'll collaborate with the Curriculum Warden to schedule skill releases, with Agent Analysts to diagnose gaps, and with the Dream Weaver team to test skills under deep-learning conditions. Every agent that performs well in your domain does so because of the blueprints you built.
What you'll do
Skill design & encoding
Decompose complex domain processes into discrete, transferable skill units
Write structured prompts, instructions, and evaluation criteria for each skill
Define edge cases, failure modes, and escalation triggers for each skill module
Version and maintain skill libraries as domain knowledge evolves
Testing & quality
Run skill agents through scenario batteries and validate output against your domain standard
Work with Bot-Psychoanalysts to review memory traces and identify skill degradation
Flag knowledge drift and submit skill update requests to the Curriculum Warden
Contribute to certification standards for agent specialization programs
Cross-functional collaboration
Partner with Agent Companions to understand real-world friction in the field
Brief Dream Weavers on which skills need reinforcement in latent learning cycles
Present skill roadmaps to the advisory board on a quarterly basis
Documentation & handoff
Maintain clear, human-readable documentation for every skill authored
Produce "skill rationale" notes explaining why decisions were encoded the way they were
Train junior Skill Architects and review their outputs
Domain specializations — pick at least one
Healthcare & clinical ops
Triage, documentation, compliance
Legal & regulatory
Contracts, risk, jurisdiction
Finance & accounting
Reporting, audit, forecasting
Engineering & infrastructure
Systems, architecture, ops
Education & research
Pedagogy, synthesis, QA
Supply chain & logistics
Procurement, routing, inventory
What we're looking for
Must-haves
5+ years of hands-on experience in your chosen domain — you've done the work, not just studied it
Ability to break down tacit knowledge into explicit, testable rules
Comfort working with AI systems — prompting, evaluating outputs, iterating
Scripting fluency in at least one of: Python, Node.js, or Bash
Strong written communication — your skill docs must be unambiguous
Nice-to-haves
Prior experience with instructional design, ontologies, or knowledge graphs
Familiarity with RAG pipelines and how retrieval affects agent behavior
Experience working in a human-AI collaborative workflow
Second domain specialization — cross-domain skills are increasingly in demand
Technical baseline
We don't expect software engineers. We expect people who aren't afraid of code. You should be able to read a script, write a simple automation, and debug a failing pipeline without needing to escalate. Fluency in at least one language is required.