Denis Vedernikov

Project: Scalable Multimodal AI Ecosystem & Enterprise Agentic Platform

Role: Lead Fullstack Architect / Backend Lead

Technical Overview:

Designed and implemented a massive, modular AI infrastructure integrating multiple modalities: LLM Orchestration, Real-time Audio/Video processing, and Generative Media (Suno, Veo, Stable Diffusion).

Key Contributions:

• Backend Architecture: Built a robust FastAPI-based service layer with sophisticated Agentic Workflows, including a custom Tool Dispatcher and Chat Orchestrator to handle complex logic between multiple AI providers.
• Multimodal Pipelines: Developed specialized services for Video Analysis, Deep Research Agents, and Audio Transcription, managed via a high-performance worker/queue system.
• Frontend Complexity: Engineered a modular Vue 3 (Vite) ecosystem with domain-driven design, implementing real-time WebSockets, custom Audio Visualizers, and a flexible Workspace for AI-driven project management.
• Infrastructure & Scalability: Integrated Redis for state management, S3 for media storage, and handled secure payment/billing systems (Fintech module).

Result: A production-ready platform capable of dynamic tool-calling and processing heterogeneous data streams in an asynchronous environment.Designed and implemented complex AI agents with RAG-based architecture and AI integrations between services. Developed and maintained backend APIs using Python and FastAPI and supported Redis and Qdrant for data and retrieval flows. Handled client and server-side development, fixed reported bugs, and formalized tasks into algorithms for subsequent coding.Full Lead Dev