AI Workforce · Module
Digital Employee Store & Role Factory
Ready-made role templates, each with its eval pack, persona policy and capability card
The Digital Employee Store is the first-party catalog of rentable role templates, built by the Role Factory from role, skill, industry, language, KPI, eval and persona packs. Every template shows what it can and cannot do, what it needs and how it was tested.
A catalog, not a blank canvas
Agent builders expect buyers to design every role, prompt and test. The Digital Employee Store, planned for W2, offers ready-made roles instead. 20 launch templates are planned, 6 of them Verified at launch: AI SDR (Inbound Lead Responder), Support Triage & Reply Assistant, Content & SEO Assistant, Collections Reminder Assistant, Meeting-to-CRM Updater and Knowledge Curator. Catalog growth toward about 100 templates in Year 2, 300 in Year 3 and 1000+ by Year 5 is a target, subject to quality gates.
The Role Factory formula
A template is Role + Industry Pack + Skill Packs + Language Pack + Tool Permissions + Knowledge Scope + KPI Pack + Eval Pack + Persona Policy. Each template version pins pack versions with semver. Skill Packs hold the Actions used, prompts or skill graphs and tool-use examples. Industry Packs add objects, vocabulary, playbooks, compliance rules and KPIs. KPI Packs define outcomes, attribution defaults and guardrail defaults, labelled estimates until data exists. Eval Packs hold golden tasks, red-team cases, simulator scenarios and pass thresholds, and cannot be weakened without review. Language Packs cover English, Hindi, Tamil and code-mixed text with Indic evals.
Honest capability cards
Every template has a capability card: what it can do, what it cannot do, autonomy and risk ceilings, required modules, eval results, cost model formula and certification level. "Cannot do" comes from the out-of-scope list and policy caps, not from marketing copy. Browsing and search work by goal, family, industry, language, channel and required apps, with entitlement badges such as "requires CRM". A recommendation from goal, stack, industry, language and budget shows a monthly investment range and return range with stated assumptions.
Quality rules
No template is published without an Eval Pack, persona policy, named owner and kill criteria. Templates without rentals are archived after a defined period. Kill-criteria monitoring watches median client ROI, margin, supervision ratio, dispute and incident rates, eval regressions and retention. Generated variants (role times industry times language times channel) inherit certification only if their own Eval Pack passes. Renting tenants get upgrade notices with a diff and eval delta and adopt by pinning a version.
For agencies and tenants
Agencies can keep private role variants visible only across their own sub-accounts, and client data stays in each client tenant. Tenants can save their own agents as private templates. A partner authoring SDK under the Marketplace Trust Standard is planned later.
Features
What Digital Employee Store & Role Factory includes
Template layer
Role templates and versions pinning pack versions, owned by One3 or partners.
Skill Packs
Actions, prompts or skill graphs and synthetic tool-use examples.
Industry Packs
Objects, vocabulary, playbooks, compliance rules and KPIs.
KPI Packs
Outcome definitions, attribution and guardrail defaults.
Eval Packs
Golden tasks, red-team cases, simulator scenarios and thresholds that cannot be weakened without review.
Language Packs
English, Hindi, Tamil and code-mixed prompts, tone and Indic evals.
Persona Policy reference
Avatar class, disclosure text per channel and jurisdiction, voice rules.
Capability cards
Can do, cannot do, ceilings, required modules, eval results, cost formula and certification.
Browse, search and recommend
By goal, industry, language, channel and required apps, with investment ranges.
Certification badges
Listed and Verified; Certified Outcome later.
Catalog quality rules
No publishing without Eval Pack, persona policy, owner and kill criteria.
Role Factory variants
Generated variants that must pass their own Eval Pack.
Kill-criteria monitoring
ROI, margin, supervision, disputes, incidents and retention per template.
Agency and tenant private templates
Variants scoped to an agency's sub-accounts or a single tenant.
Workflow
Template to listing
- 1
Draft the role with its packs
- 2
Run the Eval Pack and simulator scenarios
- 3
Generate the capability card from the spec
- 4
Check catalog quality rules
- 5
Publish as Listed, then earn Verified
- 6
Monitor kill criteria after rentals start
AI
AI skills and Digital Employees
Goal-to-role recommendation
Shortlists templates with ranges and required modules.
Capability card generation
Derives cards from the spec, including "cannot do".
Variant generation
Generates and tests role variants; failures drop to Listed.
Pack-conflict detection
Finds contradictions across pinned packs before publishing.
Digital Employees that use this module
- AI SDR — Inbound Lead Responder
- Support Triage & Reply Assistant
- Content & SEO Assistant
- Collections Reminder Assistant
- Meeting-to-CRM Updater
- Knowledge Curator
They act through the same governed Actions as people; risky steps wait for approval.
Frequently asked questions
How many templates are planned at launch?
20 launch templates, 6 of them Verified at launch. Larger catalog sizes are targets subject to quality gates.
Can I see what an agent cannot do?
Yes. Every capability card lists what the template cannot do, derived from its out-of-scope list and policy caps.
Can agencies keep their own variants?
Yes. Private role variants stay within the agency's sub-accounts and remain agency IP.
See Digital Employee Store & Role Factory in your workflow
Tell us about your team and the outcomes you want. We will walk you through the apps, the Digital Employees and the plan for your industry.