Intelligence & AI · App 17
AI Digital Employees, supervised by your team and measured honestly
Rent certified roles for a predictable monthly rent. Every agent starts in shadow mode, works only through governed Actions, and reports a monthly Investment Statement.
AI Workforce is where a business hires, supervises and measures ready-made AI Digital Employees that do real work inside its One3 apps. Each one follows the business's policies, is accountable to a named person, and reports what the business invested and what it returned, with the evidence behind every number. Digital Employees work alongside people, never instead of them, and are measured against the tenant's previous process, not against named employees.
What AI Workforce is
The public category statement for One3 is simple: rent ready-made AI Digital Employees that grow your business. Each one is trained for a specific business role, works inside your One3 software, follows your policies, is supervised by your team, and reports what you invested and what it returned, every day and every month. AI Workforce is the app that makes that possible. It is planned as 18 modules (Suite 26, the Digital Employee Economy, in the master plan) and every other app's AI plan depends on it.
Five commitments
- Work through governed Actions, not integrations. A Digital Employee acts only through Actions of modules the tenant subscribes to. Renting an AI SDR does not unlock CRM; it requires CRM. Every Action carries the agent, a run id and the accountable owner.
- Earned autonomy. Every agent starts in shadow mode (A1). Autonomy rises per skill only on evidence: agreement, edit rate, Eval Coverage Score and zero critical failures. R4 Actions never run without a human.
- Investment and return, honestly. Every agent, squad, client and agency gets an all-in cost ledger: rent, credits, pass-through, ad spend and human supervision time. Returns are labelled assisted or caused, and only holdout-validated returns may be called caused.
- Consent and disclosure are hard rules. A Digital Employee never claims to be human. Real-person likeness or voice is used only with verified, scoped, revocable consent.
- Alongside people. Agents are measured against the previous process baseline, never against named people.
How it fits together
Templates come from the Digital Employee Store, built by the Role Factory from role, industry, skill, language, KPI, eval and persona packs. A rental contract instantiates an agent in the tenant through Agents Studio, where it gets skills, knowledge scopes, a model policy and an owner. Persona Policy and the Rights Registry check every external message and render. Training Gym and simulation produce versioned releases. Actions produce outcomes such as deals, tickets and payments, and costs flow in from rent, credits, spend and supervision. Both meet in the Investment & ROI Ledger, which produces the monthly Agent Investment Statement. The Performance Arena feeds certification back into the Store. AI Governance and AI FinOps keep the whole workforce inside policy and budget.
For agencies
Agencies can rent wholesale, resell to client sub-accounts at their own price, keep private role variants as their IP, and show clients white-label statements at the agency price. Rented and agency-operated agents need a mandate issued by the client tenant.
Never promised
One3 never promises guaranteed revenue or ROI, fully autonomous growth without human owners, behaviour changes in production without a versioned release, cloning anyone without consent, or replacement of a workforce. Agents do not give legal, tax, medical or investment advice.
Problems solved
What AI Workforce fixes
Agent builders make you design every role yourself
The Store offers ready-made role templates with eval packs, persona policy and KPIs that you rent into shadow mode.
The agent talks, but someone still updates the CRM
Skills are Actions of your subscribed One3 modules (CRM, Marketing, Desk, Billing, Ads) in one tenant and one audit trail.
Nobody can tell whether AI is paying for itself
The Investment & ROI Ledger produces a monthly Agent Investment Statement with confounders, a data-quality score and a dispute window.
AI bills are unpredictable and agents stop mid-conversation
Monthly rent with itemised credits and pass-through; budget guards pause non-critical skills and hand conversations to a human queue instead.
Who is accountable when an agent does damage?
A named human owner per agent, an autonomy ladder, four kill-switch levels, forensic replay and an incident workflow.
Avatars and voices carry deepfake and consent risk
The Rights Registry blocks real-person likeness or voice without active consent, and Persona Policy enforces disclosure per channel.
Agencies want to resell AI employees under their brand
Wholesale and resale, white-label statements at the agency price and private role variants that stay agency IP.
Customers are on WhatsApp and speak Hindi, Tamil or Hinglish
Language Packs with Indic and code-mixed evals, and WhatsApp-first skills subject to policy verification.
Key capabilities
What you can do
Ready-made role templates
20 launch templates planned, 6 Verified at launch, built from role, industry, skill, language, KPI, eval and persona packs.
Shadow mode first
Every agent starts at A1, proposing while people act, and earns autonomy per skill on evidence.
Monthly Investment Statement
All-in investment and attributed return with drill-down to evidence, confounders and a dispute window.
Kill switches at four levels
Skill, agent, tenant and platform, with drills and measured stop times.
Always disclosed as AI
Persona Policy checks every external message, render and call; agents never claim to be human.
Consent for every real likeness
Verified, scoped, revocable consent for any real person's face or voice, checked at render time.
Revenue Squads
Lead Generation, Local Business Growth and Cash Collection squads with shared budgets and stop triggers.
Budget guards that degrade gracefully
Caps pause non-critical skills and route conversations to a human queue instead of failing.
Agency resale
Wholesale pools, client price lists and white-label statements at the agency price.
Modules
18 modules, one data model
Core modules come with the app. Add-ons switch on when you need them. Some modules can be bought on their own.
Agents Studio
Define, test, release and supervise every Digital Employee, skill by skill
ExploreInvestment & ROI Ledger
All-in investment and evidence-backed return for every agent, squad and client
ExploreAI Governance
A register, risk assessments, kill switches and incident handling for every agent
ExploreDigital Employee Store & Role Factory
Ready-made role templates, each with its eval pack, persona policy and capability card
ExploreAgent Rental Marketplace
Discover, trial, rent, review and return Digital Employees on clear terms
ExploreSquads & Revenue Suites
Turn a growth goal into a team of agents with one budget, KPIs and stop triggers
ExploreAvatar & Voice Rights Registry
Verified, scoped and revocable consent for any real person's face or voice
ExplorePersona Policy Engine
Every external message, render and call checked for AI disclosure before it goes out
ExploreAI FinOps
Know what every AI call costs, keep budgets in hierarchy and degrade gracefully
ExplorePrompt & Skill Library
Versioned prompts and skill definitions, compared by evals before anyone reuses them
ExplorePerformance Arena & Certification
Agent performance measured against your previous process, with earned certification
ExploreTraining Gym & Simulation
Practise with synthetic customers, improve through versioned releases only
ExploreAI Lab & Data Labeling
Datasets, labelling, experiments and models, governed for consent and licence
ExploreDigital Employees
Manage Digital Employees in your org chart, from requisition to retirement
ExploreEvaluation Studio
Golden tasks, calibrated scorers and red-team suites that gate every promotion
ExploreOutcome Contracting
Standard outcome definitions, holdouts and disputes before any outcome pricing
ExploreAgent Replacement
Swap an agent version or template without losing your context
ExploreAgent Liability Desk
Incidents, remediation, claims and insurance when an agent causes harm
ExploreWorkflows
How the work flows
End-to-end workflows inside the app and across One3.
Rent a Digital Employee
From a business goal to a supervised agent in production.
- 1.Describe the goal, industry, language, stack and budget
- 2.Get a shortlist with an investment range and stated assumptions
- 3.Run a sandbox trial on synthetic or redacted data
- 4.Sign the rental contract; missing modules are proposed first
- 5.Complete onboarding: healthy connections, knowledge scopes, persona policy, frozen baseline, named owner
- + 2 more steps →
Monthly Investment Statement
How cost and return become a statement the owner acts on.
- 1.Costs are captured daily: rent, credits, pass-through, ad spend, renders, calls and supervision time
- 2.Pass-through and spend are reconciled against provider invoices
- 3.Outcomes such as won deals, payments and resolutions are attributed under the tenant's rule version
- 4.Confounders and data quality are checked against the frozen baseline
- 5.The statement is generated with a recommendation
- + 2 more steps →
Earning autonomy
How a skill moves from shadow to supervised autonomy.
- 1.The agent proposes in shadow mode and people act
- 2.Agreement is scored per decision and per field
- 3.Eval cases accumulate across seven coverage dimensions
- 4.Promotion readiness is assessed with the missing criterion named
- 5.A person approves the promotion
- + 1 more steps →
Avatar and voice consent check
Every render with a real person's likeness is checked against consent.
- 1.The person gives verified, e-signed consent with a scope
- 2.The likeness or voice is stored as a sensitive, encrypted asset
- 3.Each render checks for active, in-scope consent
- 4.Out-of-scope renders are blocked with a reason
- 5.Revocation propagates to queued renders and scheduled posts
- + 1 more steps →
AI inside
AI skills that act through governed Actions
Every AI skill runs through the same typed Actions people use, with risk tiers, approvals and an audit trail.
Agent from description A2
Drafts a job description, skills, KPIs and at least 20 eval cases per skill; never activates what it generates.
Shadow agreement scoring A3
Scores agreement with human decisions per decision and per field to gate promotion.
Automatic demotion detection A4
Detects critical eval failures, incidents, undo spikes and loops and demotes the skill.
Statement narrative A2
Writes statement prose strictly around computed figures; models never do the arithmetic.
Confounder detection A3
Flags seasonality, discounts, price changes, existing pipeline and data gaps.
Risk assessment scoring A2
Scores skills, risk tiers, data classification and exposure; a governance admin approves.
Goal-to-role recommendation A2
Shortlists templates with an investment range and the modules each needs.
Trial scorecard A2
Summarises a sandbox trial and lists what the trial could not test.
Disclosure and impersonation check
Checks outputs for missing AI disclosure or claims of being a named human.
Digital Employees
Agents that operate AI Workforce
Rent them by the month. They work under your team's supervision and their cost and returns appear in the Investment & ROI Ledger.
AI SDR — Inbound Lead Responder
Responds to inbound leads, qualifies and books meetings through CRM, WhatsApp, email and calendar. Verified at launch.
Support Triage & Reply Assistant
Triages tickets and drafts knowledge-grounded replies in Desk. Verified at launch.
Content & SEO Assistant
Drafts briefs and content for approval; publishing stays with a person. Verified at launch.
Collections Reminder Assistant
Sends polite payment reminders through Billing, WhatsApp and email; never takes R4 Actions. Verified at launch.
Meeting-to-CRM Updater
Turns meeting notes into accurate CRM field updates. Verified at launch.
Knowledge Curator
Closes knowledge gaps and keeps answers fresh in Knowledge and Desk. Verified at launch.
Client Reporting Agent (agency)
Prepares client reports from Dashboards and the ROI Ledger, with numbers matched to the ledger.
Learn moreAI Agent Governance Auditor
Runs sample audits, policy compliance scans and drift reports over the workforce.
Integrations
Connects with the tools you already use
- WhatsApp Business Platform
- Meta Marketing API
- Google Ads API
- Gmail / Google Workspace
- Microsoft Graph
- SMS and voice with DLT
- Payment gateway (UPI links)
- Aadhaar eSign
- DigiLocker identity verification partner
- TTS and voice providers
- Microsoft Agent 365
- Salesforce Agentforce exports
- Microsoft Copilot Studio exports
- CrewAI
- n8n
- Dify
Connector availability depends on provider access and is verified before launch. How integrations work →
Security & compliance
Built to be trusted
Five-way permission intersection
An agent can act only within skills granted, its owner's permissions, tenant entitlements, per-skill autonomy and per-skill maximum risk tier.
Client mandate for rented agents
Agency-operated agents need a mandate issued by the client tenant, re-checked before every external or irreversible Action and revocable at once.
Hard limits no setting can change
An agent never approves its own proposals, never holds an R4 approval right, never manages a human and never claims to be human.
Accountable human always
Every agent has a named owner and a supervisor chain ending at a human; an agent with no accountable human pauses.
Consent evidence
Signed consent, identity verification, scope history and usage logs can be exported for disputes (verify regulatory requirements with counsel).
Cross-tenant isolation for rented agents
Templates are global, but each instance, its memory and its data stay inside the renting tenant.
Industry Editions
Industries that use AI Workforce
Frequently asked questions
Will Digital Employees replace my staff?
No. They work for your team under human supervision, are measured against your previous process, and every agent has a named owner. One3 never promises workforce replacement.
When will AI Workforce be ready?
It is being built in waves. Agents Studio, the ROI Ledger and AI Governance are planned for W1, and the Store, Squads, Rights Registry, Persona Policy and AI FinOps for W2, all inside Year 1.
Does renting an agent unlock the apps it uses?
No. An agent can only use Actions of modules you subscribe to. An AI SDR needs CRM.
How is return calculated?
The ROI Ledger attributes outcomes under your rule version, flags confounders and only calls a return "caused" when a holdout validates it. Statements have a dispute window.
Can an agent pretend to be a person?
No. Agents never claim to be human or deny being AI when asked, and avatars are always disclosed.
Can agencies resell agents?
Yes. Agencies are planned to rent wholesale, set client prices and send white-label statements at the agency price.
Is there guaranteed ROI?
No. One3 never promises guaranteed revenue, leads or ROI. Statements show evidence and confidence instead.
Works with
Related apps
Platform
One governed foundation for every human, AI agent, integration and channel
ExploreCRM & Sales
A WhatsApp-first CRM where sellers and Digital Employees work the same pipeline
ExploreMarketing
Consented, measurable marketing across WhatsApp, email, search, ads and social
ExploreCreative Studio
Brief to on-brand, rights-cleared, disclosed content with cost and return per asset
ExploreService Desk
Support on every channel, resolved on the same customer, invoice and order records
ExploreFinance
Bill, collect, pay, account, plan and comply on the records your business already keeps
ExploreIntelligence
Cited answers, one set of metrics and decisions tracked to a measured result
ExploreStrategy & Governance
Set direction, run the management rhythm and govern the company on one evidence base
ExplorePlan your AI Workforce rollout with us
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.