Two-phase approach to aggregate workforce architecture from messy sources.
Takes fragmented info (meeting notes, FigJam boards, Slack, screenshots) → produces unified workforce map.
Shows who does what, where work lives, what's unassigned/overloaded, what's automatable.
Define the buckets and categories. What types of things exist?
File: structure-schema.json
Establishes:
- Domain categories (Operations, Projects, etc.)
- Work attributes (name, frequency, criticality, etc.)
- Person types (volunteer, staff, partner)
- Gap types (unassigned, single-point-of-failure, overload)
- Automation criteria (impact/effort)
Populate with actual facts. Who, what, when.
File: lattice-schema.json
Contains:
- Actual domains with descriptions
- Work items with assignments
- People with contact info
- Detected gaps
- Automation opportunities
- Review each source doc (incubator-10-27-25.md, ai-working-group-10-27-25.md, figjam screenshots)
- Update structure schema as new categories emerge
- Agree on structure before filling lattice
- Populate lattice with actual roles/tasks/people
- Generate gap analysis (unassigned, overloaded, single-owner critical work)
- Flag automation candidates (high-frequency, low-skill, high-impact)
Domains: Functional areas (Operations, Projects, Infrastructure, Partnerships)
Work: Tasks, responsibilities, roles. Who's assigned, frequency, hours, critical?, automatable?
People: Names and types (volunteer, staff, partner, contractor)
Gaps: Unassigned critical work, single points of failure, overloaded people
Automation: AI/automation pilots with impact/effort estimates
- Critical work with one person (single point of failure)
- Person with multiple critical assignments (burnout risk)
- High-frequency manual work (automation candidate)
- Work with no owner (orphaned)
ai-working-group-10-27-25.md: AI pilot ideasincubator-10-27-25.md: Project frameworkfigjam/: Visual responsibility maps