Blog
Apache Kafka Part 1: What Stream Processing Is and Why It Changes Everything
Kafka is not a typical message broker — it’s the distributed nervous system powering Netflix, LinkedIn, and Uber. It handles millions of events per second without losing a single one, with guaranteed ordering per partition. This first installment explains the core concepts (topics, partitions, offsets, consumer groups) using a real use case: the 50 ARPA Piedmont stations from the Smart City project at Politecnico di Torino.
Supabase: the Open-Source Backend for Your Vibe-Coded Apps
Lovable and Bolt build the frontend in minutes. But where does user data live? How does login work? Who can see what? Supabase answers all of these questions: managed PostgreSQL, ready-to-use authentication, file storage, and Row Level Security — all free up to a generous limit, all integrable in a single click from the main vibe coding tools.
Sentiment Analysis & Topic Modeling: What Your Customers Really Mean
You have 200 reviews, 500 support tickets, 1,000 social media comments. Reading them all would take days — and you’d still miss the most important patterns. Sentiment Analysis and Topic Modeling solve exactly this: in ten minutes you get the emotional tone of every text, recurring themes grouped automatically, and a strategic summary that manual reading would never have produced.
Multimodal AI: Analyze PDFs, Images and Documents with Claude, GPT-4 and Gemini
AI no longer reads only text. Claude summarizes a 10-page quote in 30 seconds. GPT-4 Vision transcribes data from a dashboard screenshot into a ready-to-use table. Gemini 1.5 Pro navigates 1,000-page documents citing the sources. This guide shows how they work, when to use which tool, and where the time savings are measurable — with real screenshots from live sessions.
RAG: How to Build a Chatbot That Actually Knows Your Company
RAG (Retrieval-Augmented Generation) is the technique that transforms a generic LLM into an assistant that answers directly from your internal documents. This guide shows how the pipeline works — chunking, embedding, vector store, retrieval — and how to implement it today: without code using Claude Projects and Chatbase, or with a custom build using LangChain and LlamaIndex.
Grafana: Professional Dashboards for Technical and IoT Data
Grafana isn’t a replacement for Looker Studio — it’s a fundamentally different tool for a different problem. Where Looker Studio excels with business reports for stakeholders, Grafana is the visualization layer built for real-time technical data: IoT metrics, time series, infrastructure monitoring. It connects to InfluxDB, PostgreSQL, Elasticsearch, and dozens of other data sources. Free, open source, with native alerting built in.
