dataaaaa!

Workplace AI Omni is a self-hosted AI assistant that unifies workplace data through BM25 and pgvector search. It uses Postgres for all storage, eliminating the need for separate vector databases.

is a 200-line Python script that implements a complete GPT training and inference pipeline without dependencies. It includes a custom autograd engine, tokenizer, and Adam optimizer to demonstrate the core mechanics of LLMs. Everything else is just efficiency.

Bonnard is an agent-native semantic layer providing a unified schema for MCP servers, SDKs, and markdown dashboards. It ensures metric consistency across AI agents and internal tools. One set of metric definitions, every consumer gets the same governed answer.

is a Rust-powered engine designed for high-performance RAG pipelines. It processes gigabytes of data from formats like CSV and PDF in seconds using O(1) memory. 40x faster than LangChain with O(1) memory usage. It supports smart splitting strategies and HTML cleaning.

is an open-source time-series database built on PostgreSQL and TimescaleDB. It uses a three-dimensional temporal model to manage overlapping forecast revisions and auditable updates.

is a VS Code extension that transforms SQL queries into interactive execution flow diagrams. It visualizes complex logic using color-coded nodes for joins, aggregates, and CTEs.

Flowrs is a TUI application for Apache Airflow, enabling terminal-based monitoring, inspection, and management of Airflow DAGs.

This lab offers hands-on practice for the Databricks Certified Data Engineer Associate exam, covering Auto Loader, Medallion Architecture, SCD Type 2, Unity Catalog, and Databricks Jobs orchestration.

Atomic Agents is a lightweight, modular framework for building AI pipelines using "atomicity."

The technical challenge involves parsing a 100 million row CSV dataset into JSON using PHP, with specific formatting rules and data validation requirements.

DeepAnalyze is an open-source agentic LLM for autonomous data science, supporting the entire data pipeline from preparation to report generation, with CLI, WebUI, and JupyterUI interfaces.

Hugging Face Skills provides a standardized format for defining AI/ML tasks such as dataset creation, model training, and evaluation, compatible with tools like OpenAI Codex, Claude Code, Gemini CLI, and Cursor.

Ducktape is a lightweight REST API for DuckDB, leveraging HTTP/2 for NDJSON streaming, optimizing data ingestion and queries with near-native performance.

This GitHub repository provides code examples for spatial data management using DuckDB, covering SQL, Python, and advanced geospatial analytics.

Structured CIA World Factbook archive with 1M+ parsed data fields, offering advanced analytics, custom queries, and interactive visualizations for geopolitical exploration.

Comprehensive timeline of language models, from Transformer (2017) to GPT-4, Claude, Gemini, LLaMA, Mistral, and DeepSeek, powered by Splox visualization.

DuckDB.ExtensionKit enables the development of native DuckDB extensions in C# using .NET AOT compilation, with code generators to streamline integration.

adbcli is a Rust-based command-line tool for querying databases via ADBC, featuring multi-database support, interactive SQL shell, and export of results to JSON, CSV, or Arrow IPC files.

DataStudio is a browser-based Python/SQL notebook leveraging WebAssembly, DuckDB for analytical processing, and Pyodide for Python, with local data management.

Datui is a high-performance terminal UI for exploring and analyzing tabular data, supporting Parquet, CSV, JSON, and other formats, with features like SQL filtering and data visualization.

OptimizeQL is an AI-powered SQL query optimizer leveraging LLMs to analyze execution plans and suggest improvements such as index creation or query rewrites.

IronClaw is a secure AI agent framework built in Rust, featuring a Zero Trust architecture, a 13-step security pipeline, and supporting 25+ LLM providers and 20+ communication channels.

NTransformer is a high-efficiency C++/CUDA LLM inference engine designed to run models like Llama 70B on an RTX 3090, leveraging a 3-tier adaptive caching system (VRAM, RAM, NVMe) and streaming techniques to optimize performance.

Yardstick is a DuckDB extension implementing SQL measures to simplify percentage calculations and time comparisons without CTEs or window functions.

MirrorMetrics is a benchmarking tool for evaluating Stable Diffusion LoRA models, leveraging InsightFace for local biometric analysis and generating an interactive dashboard with Plotly.