Revitk
AI-Powered Sales Enablement & GTM Intelligence Platform
Revitk is a multi-tenant B2B SaaS AI platform that enables sales consultancies and enterprise teams to operationalize proprietary sales methodologies at scale. The platform allows organizations to load customer-specific GTM (Go-To-Market) frameworks, knowledge bases, and sales methodologies into AI—delivering both structured sales deliverables and free-form contextual chat through a dual-mode AI architecture.
The key innovation is a dual-mode AI system: - Generator Mode for ontology-driven, structured outputs (discovery scripts, business cases, VSLs, email sequences) - Chat Mode for conversational exploration with full customer and GTM context loaded - Sales Coach Mode for AI-powered call transcript analysis, scoring reps against sales methodologies, and tracking performance over time
This enables sales teams to generate precisely formatted outputs, freely explore ideas without losing organizational context, and continuously improve through data-driven coaching insights.
Tech Stack
Project Gallery
Main dashboard with customer overview and quick actions
Dual-mode AI chat with GTM context injection and skill selection
Real-time markdown preview for generated deliverables
PDF document generation and preview
GTM framework management with RDF/TTL ontology
Team and rep performance tracking with competency scores
Organization analytics with token usage and skills breakdown
Super-admin client organization management
Key Responsibilities
- Architected a dual-mode AI system supporting both structured generators and contextual chat via a unified skill engine
- Designed a dynamic context injection system loading customer-specific GTM frameworks (personas, pain points, competitors) into AI workflows
- Built a skill execution engine with real-time SSE streaming and Redis-based status updates
- Developed the Sales Coach skill for AI-powered call transcript analysis, scoring reps against sales methodologies with automatic competency extraction
- Implemented a Performance Dashboard tracking team and individual rep scores across four competencies with rolling averages and trend indicators
- Built coachability tracking system that identifies coaching focus areas and tracks whether reps attempt improvements
- Developed multi-tenant architecture with strict organization-level data isolation and platform-level admin (Thanos Mode) capabilities
- Implemented role-based access control (RBAC) with Super Admin, Org Owner, Org Admin, and Member roles across platform and organization levels
- Built Admin Client Organizations panel for super-admins to view all client data, chat history, analytics, and performance metrics
- Designed GTM framework storage using RDF/TTL ontology format with structured fields
- Implemented three-tier caching (Redis → PostgreSQL → MinIO) for fast AI context retrieval
- Developed real-time AI streaming chat with resumable streams for network resilience
- Built AI-powered document generators producing PDF, DOCX, Excel, PPTX, and Markdown deliverables
- Implemented team management with email invitations, role updates, and member lifecycle management
- Built organization analytics dashboard tracking token usage, model breakdown, and skills usage statistics
Key Achievements
- Enabled non-technical skill iteration by separating ontology ownership from output refinement
- Reduced GTM context load time by approximately 40% using Redis caching with intelligent invalidation
- Achieved 99.9% streaming reliability with resumable stream architecture
- Reduced manual sales deliverable creation time by approximately 80%
- Implemented dual-mode AI architecture allowing seamless switching between structured generators and free-form chat
- Built Sales Coach skill that automatically classifies call types (discovery, demo, proposal, appointment setting, one-call close) and scores against appropriate frameworks
- Developed competency scoring system tracking four key areas: Process Control, Discovery Depth, Objection Handling, and Outcome Positioning
- Implemented discovery sub-skills assessment (current state, desired state, gap analysis, root cause, decision process, objections handling)
- Built coachability tracking that identifies the lowest-scoring area for each rep and monitors improvement attempts across calls
- Created Thanos Mode allowing platform admins to access all client GTM frameworks simultaneously for cross-organization insights
- Implemented web search, web fetch, and document generation tools with automatic tool selection
- Built real-time skill execution with SSE streaming of tool lifecycle events and progress tracking
- Developed read-only chat view for admin oversight of client conversations
- Multi-Format Document Generation & Preview: Enabled generation of PDF, DOCX, Excel, PPTX, and Markdown deliverables with real-time previews
- Built intelligent chat system that automatically injects organization-specific GTM data into AI context
- Implemented Most Improved rep tracking and team-level coaching focus identification
