Projects by Ilir Ivezaj

Data Platform

Enterprise Data Integration & Analytics Platform

Architected and built an enterprise data integration platform processing millions of records daily across Oracle, SQL Server, PostgreSQL, MySQL, MongoDB, and Snowflake. Includes automated schema detection, real-time synchronization, multi-tenant analytics, KPI dashboards, and comprehensive audit logging.

Role: Lead Architect & Product Engineer

.NET Angular Azure Oracle Snowflake

Compliance

Regulatory Compliance & Supply Chain Platform

Built a comprehensive regulatory compliance platform with real-time product verification, exceptions management, automated audit trails, and 6-year data retention. Serves enterprise clients across pharmaceutical, healthcare, and manufacturing sectors with configurable track-and-trace capabilities.

Role: Product Engineer & Conference Speaker

.NET Angular SignalR Azure

AI / Multi-Agent

Multi-Agent AI Development Framework

Designed and built a production multi-agent AI development environment orchestrating Claude, GPT, Gemini, and Cursor agents. Features shared MCP memory server, coordinated task execution, safety guardrails, and 3-5x engineering productivity gains across architecture, implementation, testing, and review.

Role: Creator & Architect

Python TypeScript MCP Protocol LLM APIs

GPU / ML

GPU-Accelerated Analytics & Local LLM Platform

Built a GPU-accelerated machine learning platform on RTX 5080 with CUDA 13.1. Runs local LLM inference (Qwen 3.5-35B) via llama.cpp, trains custom classifiers with PyTorch, and deploys real-time inference APIs with FastAPI. Zero API costs, complete data privacy, sub-second response times.

Role: Engineer & Researcher

PyTorch CUDA llama.cpp FastAPI

AI Automation

AI Document Intelligence Engine

Built an AI-powered document processing engine that extracts structured data from invoices, contracts, and forms using OCR, NLP text classification, and custom trained models. Automates data entry, reduces processing time by 80%, and routes documents to correct workflows automatically.

Role: AI Engineer

Python Hugging Face FastAPI Azure AI

Predictive

Predictive Analytics & Anomaly Detection System

Designed an enterprise analytics system that forecasts order volumes, detects suspicious patterns for regulatory compliance, predicts transaction failures before they occur, and identifies operational bottlenecks. Trained on millions of historical records with continuous model retraining.

Role: ML Engineer & Architect

scikit-learn PyTorch Snowflake Power BI

Startup / SaaS

Albahub — Workflow Automation Platform

Founded and built Albahub, a workflow automation platform that helps operationally complex businesses streamline their operations. Connects CRMs, ERPs, scheduling tools, accounting software, and messaging platforms to automate intake, handoffs, documentation, approvals, billing, and communications. AI-powered document intelligence, visual inspection, and predictive analytics.

Role: Founder & CTO · beta.albahub.io

React .NET FastAPI Oracle Cloud Auth0 Cloudflare

SaaS

Business Intelligence Dashboard

Designed and built a SaaS business intelligence platform featuring interactive dashboards, automated reporting, CRM integration, and real-time analytics. Serves businesses with actionable insights through customizable visualizations and data-driven workflows.

Role: Founder & Full-Stack Engineer

Next.js Python PostgreSQL HubSpot API

AI / ML

AI-Powered Analytics Engine

Built a GPU-accelerated machine learning system for predictive analytics and anomaly detection across large-scale datasets. Features automated model training, real-time inference APIs, and integration with business intelligence platforms for actionable insights.

Role: Lead Engineer

PyTorch CUDA FastAPI Redis

Developer Tools

Multi-Agent Development Framework

Designed an AI-powered multi-agent development framework that orchestrates multiple AI models for software engineering tasks. Features coordinated code generation, automated testing, and intelligent code review across complex codebases.

Role: Creator & Maintainer

Python TypeScript LLM APIs MCP

Infrastructure

Distributed Event Processing System

Architected a high-throughput event streaming platform deployed on Kubernetes, handling hundreds of thousands of events per second. Features automatic scaling, dead-letter queues, exactly-once processing semantics, and comprehensive observability.

Kubernetes Kafka Go Prometheus

Automation

Infrastructure Automation Toolkit

Created a comprehensive infrastructure automation toolkit for provisioning, monitoring, and managing cloud resources across multi-cloud environments. Includes Terraform modules, monitoring dashboards, cost optimization scripts, and security audit automation.

Terraform Python Bash Grafana

Enterprise Platforms