Context: Evolution from managing federal-scale divisions (150+ people) to building SaaS products. Full-cycle development: from operational models and BPMN design to production code and monetization.
Role
Design and development of scalable internal systems and SaaS solutions with complex business logic. Application architecture integrating AI layers (LLM, RAG) and payment systems (Stripe), focused on reducing operational costs through automation.
Responsibilities
- Data Architecture: Database structure design (PostgreSQL), vector storage (Pinecone/Supabase/pgvector) for AI modules, complex role models (RBAC) with granular access control
- AI-assisted coding: Claude Code (VS Code), development of production systems with LLM in a loop. Development of LLM-based AI consultants/agents (OpenAI API, LangChain/LangGraph) using RAG (Retrieval-Augmented Generation) and vector knowledge bases for specialized protocols
- Fintech & API: Stripe integration (Payments & Subscriptions), recurring payment setup; VoIP telephony integration and data exchange with external CRMs
- Workflow Automation: Building end-to-end automation (n8n, custom API solutions), transforming fragmented operations into manageable digital workflows
- Product Development: MVP/PoC creation (mini-CRM with booking, task management), requirements formalization (BRD, SRS), product management through to monetization stage
Results
- Development using Claude Code — from architecture to production deployment
- CRM Launch with AI Consultant for an international wellness company: 60% reduction in manual operations, automated service booking and warehouse/inventory reservation
- Subscription Model Implementation: Stripe integration for recurring payments and automated subscription management
- Enhanced Control: Transition from Excel and manual control to system workflows with checkpoints and real-time notifications
- Scalability: Solution architecture allows adding modules (new integrations, AI features) without core system re-engineering
