- Home
- AI Data Analysis Tools
- ADA: Automated Data Analyst
On the recordRead 7 Oct 202621 facts
ADA: Automated Data Analyst
On the record
- AI safeguards
- ADA validates model output against a typed schema and the actual data, and it never executes model-generated code.github.com · read 7 Oct 2026
- Anomaly detection
- Anomaly flags identify periods outside a calibrated band, sized so a stable series false-alarms about once in twenty analyses.github.com · read 7 Oct 2026
- Ask ADA
- Plain-English questions are answered with pandas calculations, and the calculation is shown with the answer.github.com · read 7 Oct 2026
- Calculations
- ADA shows the calculation behind its findings and labels recommendations as interpretation rather than cause.github.com · read 7 Oct 2026
- Dashboard
- The dashboard includes trends, segment breakdowns, a movement waterfall, and a segment-by-period heatmap.github.com · read 7 Oct 2026
- Downloads
- ADA can download a Markdown executive brief and a cleaned CSV.github.com · read 7 Oct 2026
- Exports
- Users can download a Markdown executive brief and a cleaned CSV.github.com · read 7 Oct 2026
- File limits
- The stated limits are 25 MB per file and 250,000 analyzed rows; supported formats are .csv, .xlsx, and .xlsm.github.com · read 7 Oct 2026
- Forecasting
- Forecasting uses a guarded baseline that declines to run on thin history and reports when it does no better than assuming no change.github.com · read 7 Oct 2026
- Hosted demo privacy
- The hosted demo holds uploaded files in session memory and does not write them to a database.github.com · read 7 Oct 2026
- Integration
- The maker describes an optional OpenAI integration for planning questions the rule parser cannot handle; pandas executes the resulting calculation locally.saineshnakra.medium.com · read 7 Oct 2026
- License
- ADA is MIT licensed, and the README says it may be used, changed, and shipped commercially with the copyright notice retained.github.com · read 7 Oct 2026
- Local processing
- When self-hosted, file parsing, cleaning, charts, forecasting, and supported question answering run locally with pandas and no network call.github.com · read 7 Oct 2026
- Local use
- ADA can be run as a Streamlit app from its source repository, and the README says no API key is required.github.com · read 7 Oct 2026
- Maker
- The repository credits Sainesh Nakra as the maker, and his site lists San Jose, California as his location.github.com · read 7 Oct 2026
- Optional AI
- Optional, button-triggered AI calls require a key and receive schema and computed evidence rather than uploaded rows; the README notes evidence may include segment labels.github.com · read 7 Oct 2026
- Privacy
- When run locally without an API key, file parsing, cleaning, schema detection, charts, anomaly detection, forecasting, and Ask ADA are computed locally.github.com · read 7 Oct 2026
- Purpose
- ADA analyzes CSV and Excel files to create a dashboard, executive brief, anomaly flags, forecast, and answers to plain-English questions with calculations shown.github.com · read 7 Oct 2026
