Ivan T.

🤖 I build production AI products for B2B companies: AI agents, LLM/RAG systems, and full-stack SaaS platforms. The hard part is not the demo; it is making the product reliable for real users, workflows, data, permissions, and integrations.

📊 20+ AI products delivered | 30+ software products and systems launched | OpenAI Select Partner

I work as a hands-on AI Developer and Python Full Stack Developer, taking end-to-end ownership of scope, architecture, FastAPI backend development, agentic AI workflows, API integrations, data, evaluation, and production launch.

You work directly with me. When a project needs more capacity, I lead a small Unrealos engineering team while remaining accountable for architecture, delivery, and communication.

🎯 I am a strong fit when you need to:

• Build an AI SaaS product or MVP from idea to launch
• Develop AI agents, RAG, hybrid search, scoring, or decision logic
• Add generative AI capabilities to an existing web application
• Turn an AI prototype into a production system with real business rules and integrations
• Rescue or extend a SaaS platform constrained by backend, data, permissions, or reliability problems

🚀 Selected recent work:

• Generative AI Template Platform - multi-agent prompt-to-template generation and editing, structured JSON validation, model evaluation, and production editor integration
• Multi-Agent Real Estate Platform - separate buyer, seller, and new-build agents with hybrid SQL/vector search and CRM escalation
• HR AI Document Processing Platform - Claude-powered CV and certificate extraction with human review, permissions, and live staffing workflows
• B2B AI Sales Intelligence Platform - company research, opportunity scoring, and AI-generated sales preparation for enterprise deals
• AI Tender Evaluation System - weighted scoring and RAG-based document compliance validation

⚙️ Core stack:

• Python, FastAPI, React, REST APIs, WebSockets
• OpenAI API, Claude API, Gemini, Perplexity, LangChain, LangGraph, Langfuse
• Supabase, PostgreSQL, pgvector, vector databases, Firebase/Firestore
• Structured outputs, RAG, hybrid search, evaluation, observability, permissions, and third-party integrations