Platform Features
Built for the future of work
Every feature is designed around one goal: let your team focus on high-value work while AI and remote professionals handle the rest.
AI Workforce
Multi-agent AI that thinks before it acts
A Worker Agent proposes, a Manager Agent reviews. Only safe, confident proposals auto-execute. Everything else goes to your team.
Dual-Agent Pipeline
Worker Agent (GPT-4o) generates structured action proposals. Manager Agent (Claude) scores risk, compliance, accuracy, and completeness.
Human-in-the-Loop (HITL)
High-risk actions always escalate to humans. Your team approves or rejects with mandatory feedback — full audit trail.
Auto-Execution Rules
Confidence ≥ 90% + low risk = auto-execute. Emails and escalations are hard-coded to always require human review.
Industry-Specific AI Personas
Each workspace gets a tailored AI persona — CS rep, sales dev rep, virtual assistant, or project manager — with domain-specific prompts.
Structured Output (Zod)
All AI outputs are constrained by typed schemas. No free-form text, no hallucinated actions — just validated, predictable proposals.
Multi-Provider Architecture
OpenAI GPT-4o drafts the action, Anthropic Claude reviews it for risk and confidence. Each model handles what it does best.
Industry Workspaces
One platform, four industry modes
Each workspace adapts its task board, starter data, and AI behaviour to the vertical — no configuration needed.
Customer Service
Board: Ticket Table
Support ticket management with priority, status, assigned agent, and SLA tracking. AI persona acts as a CS representative.
Sales
Board: Kanban Pipeline
Leads → In Progress → Proposal → Closed Won pipeline. AI persona acts as a sales development representative.
Virtual Assistant
Board: Task List
Flexible task list for administrative work, scheduling, research, and coordination. AI persona acts as a virtual assistant.
Project Management
Board: Timeline View
Phase-based timeline with status indicators and milestone tracking. AI persona acts as a project manager.
HR Management
Talent Profile & Management System
Talent builds professional profiles, companies browse and hire, and VcrewX manages the entire workforce — all from one platform.
Talent Onboarding
Talent self-registers, builds a professional profile, and becomes visible to companies once their profile reaches 70% completeness.
Upwork-Style Profiles
Skills, work experience, education, certifications, languages, portfolio, hourly rate, and availability — all editable by the talent themselves.
Document Management
Upload and store contracts, tax forms, IDs, and onboarding docs in Firebase Storage. PDF, JPEG, PNG — max 10 MB.
Profile Completeness Scoring
10-point scoring system automatically tracks profile progress. Talent is promoted from applicant to available at 70%+ completeness.
Talent-Company Matching
VcrewX assigns available talent to company workspaces. Both the contractor profile and workspace are updated atomically.
Talent Directory
Searchable directory with status, skills, availability, experience level, and assigned company. Live Firestore data.
Real-Time Dashboards
Live data, everywhere
Every portal gets real-time updates from Firestore. Task changes, AI proposals, and team activity appear instantly.
Platform Dashboard
Active companies, talent pool metrics, workspace count, pending AI approvals, and task completion metrics — all managed by VcrewX.
Company Dashboard
Company overview, hired contractors with names, task count with completion rate, and workspace status badge.
Talent Dashboard
Pre-hire: profile completeness tracker and checklist. Post-hire: task board, timesheet, and assignment details.
AI Activity Monitor
Dedicated monitoring dashboard: total invocations, tokens consumed, average duration. Detailed log table per AI call.
Live Notifications
Bell icon shows pending AI approvals (company) or assigned tasks (talent) with confidence scores and deep links.
Task Board Views
Industry-adaptive: tickets (CS), Kanban (Sales), list (VA), timeline (PM). All powered by Firestore onSnapshot.
Security
Enterprise-grade from day one
Multiple layers of protection for your data, your users, and your AI.
Role-Based Access Control
Multiple security roles with two-layer route protection: Edge middleware + client-side RoleGuard. Every user sees only what they should.
Server-Side Auth Checks
Every server action verifies caller identity via httpOnly session cookies and Firestore workspace membership. Cryptographically secure.
Firebase App Check
reCAPTCHA Enterprise attestation prevents direct API abuse. Auto-refreshing tokens protect every Firebase call.
Zod Input Validation
Every server action validates all inputs through Zod schemas. Document IDs restricted to alphanumeric + underscores.
Prompt Injection Defense
15+ regex patterns detect injection techniques. Content sandboxing wraps external data. Output schemas constrain AI responses.
File Upload Security
Whitelist-only file types (PDF, JPEG, PNG), 10 MB cap. Firebase Storage rules control access per user role.
See it in action
Ready to explore? Sign up and experience the full platform.
Looking for work? Create your profile