Tried it, wrote it down.

What I learned running websites, APIs, databases and small experiments on an 8 GB Core i5 laptop, managed from Windows.

The setup I use: Ubuntu Server, SSH from Windows, Docker, Cloudflare Tunnel, and a Cockpit dashboard with battery and display pages.

Explore the live JobLab dashboard, then inspect the Python, Docker, Terraform, ECS Fargate, S3, and Supabase pipeline that rebuilds its optimization-focused dataset each day.

Inside JobPilot's SvelteKit, Convex, FastAPI, Supabase, Gmail, and task-agent architecture, from a personal search request to an updated application board.

How I deployed a shared MLflow tracking server with PostgreSQL metadata and external artifacts, then connected JobLab tracing, evaluation, feedback, and prompt optimization to it.

How I containerized a FastAPI optimization API, deployed it from GitHub to Railway, exposed /solve, and connected a separate frontend to the public endpoint.

How I used a five-year supply-chain network demo to test CAPEX, fixed OPEX, transport cost, opening time, and capacity tradeoffs before building the optimizer.

How I built the Next.js, TypeScript, Mantine, ECharts, Zustand, and Supabase frontend that turns JobLab's normalized job records into interactive analysis.

How I connected the oploy.eu Django and Wagtail site to Railway PostgreSQL, configured SSL and persistent connections, checked latency, and monitored the real deployment.

A genetic algorithm picks reorder points for five products. When I re-measured its best answers with 200 runs, they were 11 to 15 percent worse than it said.

A multi-product (r,Q) and (s,S) inventory simulation, what a single 60-period run hides, and a fill-rate bug I only found when I reran it.

From my Master's thesis, published in 2025: an ANN metamodel against Kriging for simulations that take minutes per run, and where the comparison stops.

A schedule from my repair-shop scheduling work, with the model objectives, constraints, solution methods, and source code behind it.