Compare your greenhouse records. Know what needs checking.
Bring climate measurements and optional irrigation history. Map your columns, choose a time window, and get a report that keeps calculations, source evidence, and missing information visible.
The evidence stays with the answer.
- 01
What the records establish
Compare matching periods by site and zone. Keep source filenames, row references, timestamps, and units attached to the result.
- 02
What was calculated
Review time-weighted climate summaries, time within your targets, VPD, dew point, and observed light totals with their assumptions.
- 03
What remains unknown
See rejected rows, gaps, partial light coverage, and missing irrigation history. Logged controller activity does not establish delivered water or soil moisture.
- 04
What needs checking
Use an evidence-backed inspection brief to choose the next measurement or physical check, then export the report as JSON and Markdown.
CSV and JSON, with explicit mapping.
Use UTF-8 CSV with unique headers or a JSON array of row objects. Supply a timestamp with a UTC offset, site, zone, and at least one supported environmental measurement. Choose the source columns and units, then confirm the mapping before analysis.
- Air temperature, relative humidity, PPFD, CO₂, and optional leaf temperature.
- Optional irrigation event start, zone, and logged duration. Flow estimates require an explicit flow assumption.
- Maximum 5 MiB and 20,000 rows per file. Timestamps must include Z or a UTC offset.
- This release supports CSV and JSON. XLSX, PDF extraction, automatic device connections, and lux-to-PPFD conversion are unavailable.
Synthetic examples
Try the complete two-week workflow.
These files are generated examples. They do not describe a real farm or a live sensor feed. Load them through the same mapping and analysis steps used for your records.
Illustrative result · light integration
21.6 mol/m² observed light
Constant PPFD of 500 µmol/m²/s over 12 completely observed hours integrates to 21.6 mol/m² for that interval. It is not a complete-day DLI. Gaps reduce coverage and remain visible rather than being filled without a bound.
Analysis runs in your browser.
The web application processes selected files in memory and does not send their contents to Streamline. Clear session releases the application's references; originals and downloaded reports remain on your device. Opening the site still makes normal hosting requests. Read the data-handling notice for the separate rules that apply when using ChatGPT or Codex.
The release analyzes supplied records. It does not control equipment, continuously monitor a greenhouse, or diagnose disease from ambient readings. Use the report to decide what to measure and inspect next. For help, contact bennett@streamlinefarms.com.