DataVeritas

Live demo

DataVeritas is a Streamlit dashboard for profiling tabular datasets and auditing data quality.

Install and run

DataVeritas can be installed from PyPI once a release has been published:

python -m pip install dataveritas
dataveritas

The dataveritas command starts the dashboard and opens it in your browser. The explicit form is also supported:

dataveritas serve

Streamlit server options can be passed through:

dataveritas serve --server.port 8502

The dashboard runs locally; it does not require a separate API server. The AI assistant tries Groq first using GROQ_API_KEY and falls back to local Ollama if Groq is unavailable. API keys are not included in the pip package; each user provides their own key in their environment or Streamlit secrets. For the fallback, install and start Ollama, then download the default model:

ollama pull llama3.2

Set OLLAMA_MODEL to use a different locally installed model, or OLLAMA_HOST if Ollama is running at a non-default address. If neither provider is available, the dashboard explains what needs to be configured.

Install from a checkout

python -m pip install .
dataveritas

For development, install the declared dependencies and run Streamlit directly:

python -m pip install -r requirements.txt
streamlit run dashboard.py

Package layout

• dashboard.py — Streamlit user interface
• sections/ — dashboard views and interactions
• core/ — data quality analysis and supporting logic
• dataveritas/cli.py — command-line launcher installed as dataveritas