Running a data management office as software, with AI help attached to the people doing the work instead of parked in another portal. Humans own every decision, and it runs entirely inside the organization's own network.
A model is fluent everywhere and grounded nowhere. Syaq sits behind whichever model you already use and hands it the Saudi part, dated and traceable to a real source. Arabic governs, and it stays free.
From binary to assembly to Python to AI agents. Tracing 80 years of abstraction and what happens when code writes itself.
One platform for both halves of an analytics stack: Superset dashboards and Dremio self-service on one side, document routing, RAG chat, and quality profiling on the other. The models run locally on Ollama, so nothing leaves the box.
Batch and streaming pipelines running side by side over the same live market feed: Airflow into DuckDB on one path, Kafka into Spark on the other, Superset on top. Open source, 73 tests, one docker-compose up.
An end-to-end pipeline on SQL Server 2022: REST and CSV ingestion, four layers from Raw to Analytics, and dashboards that count every row received, stored, and rejected. Airflow orchestrates it, Docker Compose runs it.
Production-grade ELT pipeline with asset-based orchestration, SQL transformations, and automated data quality testing.
Step-by-step guide to implementing medallion architecture for NYC Taxi data using modern data engineering tools.