Skip to content

No account required

Start a project

Your data. Finished.

Data Analysis App turns messy files into clean data, a checked analysis, a dashboard and a deck. No prompting required.

See every file type we accept →
  • .xlsx. Excel workbook. Every sheet read, including the ones that are notes rather than data.
  • .xls. Excel 97–2003. Read the same as a modern workbook.
  • .xlsm. Excel, macro-enabled. Values and formulas are read. Macros are never executed.
  • .ods. OpenDocument spreadsheet. LibreOffice and OpenOffice workbooks, read like Excel.
  • .numbers. Apple Numbers. Tables are pulled out of the Numbers package sheet by sheet.
  • .csv. CSV. Delimiter, quoting and encoding detected rather than assumed.
  • .tsv. Tab-separated. Read as delimited text.
  • .pdf. PDF. Checked for a text layer first. A PDF written by software keeps its table structure; a scan is read as an image and told apart from the first kind.
  • .docx. Word document. Tables are lifted out of the document; the prose around them is kept as context.
  • .pptx. PowerPoint deck. Tables and the data behind native charts are read slide by slide.
  • .rtf. Rich text. Tables extracted, formatting discarded.
  • .json. JSON. Nested objects are flattened into tables, keeping the path as the column name.
  • .ndjson. Newline-delimited JSON. Streamed a record at a time, so an export of any size is fine.
  • .xml. XML. Repeated elements become rows; attributes become columns.
  • .parquet. Parquet. Read columnwise, with the file's own types kept.
  • .sqlite. SQLite database. Every table read, with the foreign keys used to work out how they relate.
  • .db. SQLite database. Read as SQLite when the file is one, and reported plainly when it is not.
  • .pbix. Power BI report. The tables inside the report are read out of its model, along with the measures that were defined on top of them, so your own definitions are used rather than re-invented.
  • .pbit. Power BI template. A template carries the model and the measures without the data, and is read for both.
  • .twbx. Tableau packaged workbook. The packaged extract is read as data, and the workbook beside it supplies the calculated fields.
  • .twb. Tableau workbook. Calculated fields and the shape of each sheet are read.
  • .hyper. Tableau extract. Read straight through, table by table.
  • .ofx. OFX statement. Transactions come with their own type codes and references, which makes matching far better than a CSV.
  • .qfx. Quicken statement. Read as OFX, which is what it is underneath.
  • .qbo. QuickBooks statement. Read as OFX, with the QuickBooks account mapping kept.
  • .sta. MT940 statement. The SWIFT tags are parsed into transactions, balances and references.
  • .sav. SPSS data. Variable labels, value labels and missing-value codes are all read, so a 5 stays “Strongly agree” and a 99 stays missing.
  • .zsav. SPSS data, compressed. Read the same as .sav.
  • .por. SPSS portable. Read the same as .sav.
  • .dta. Stata data. Variable and value labels are kept, so the codebook does not have to be re-typed.
  • .sas7bdat. SAS data. Read with its formats, including dates.
  • .xpt. SAS transport. The interchange format SAS itself exports. Read like a .sas7bdat.
  • .png. PNG image. Screenshots of a dashboard or a table are read, and every figure is checked against the totals in the same image.
  • .jpg. JPEG photo. A photograph of a printed report or a handwritten sheet is read into a table.
  • .jpeg. JPEG photo. The same as .jpg.
  • .heic. iPhone photo. Accepted as it comes off the phone. Safari usually hands it over as a JPEG.
  • .webp. WebP image. Read like any other screenshot.
  • .tiff. TIFF scan. Multi-page scans are read page by page.
  • .txt. Plain text. Read as notes, or as data when the contents turn out to be a table.
  • .md. Markdown. Markdown tables become tables; the rest becomes context for the analysis.
  • .zip. Zip archive. Unpacked, and each file inside is treated as though you had uploaded it yourself.
01

Give us the mess.

Excel exports. PDFs. Three versions of the same file. Column names nobody chose on purpose. Duplicate rows, subtotals stranded in the middle of the data, dates in four formats.

You do not have to tidy any of it first. That was the part you were dreading.

ACCEPTSXLSX, CSV, PDF, JSON, PBIX and ZIP/ROWS 7,219/REPAIRS 56
READING YOUR FILES

Free your data from documents.

Five passes turn a document into data: inside the first of the six stages every project runs. Watch one work.

SELECT A STEP TO SEE IT RUN
ACME DISTRIBUTION / Q4 STATEMENT
TABLE 5×4
West
16,550
18,775
18,950
Central
5,125
5,600
5,675
READING LAYOUT…
WITHOUT PROMPTS

It doesn’t wait for you to ask the right question.

Most work waits for you to know what to ask for. This reads every file, forms its own hypotheses, tests them, and eliminates the ones that are inconvenient for its story.

H1Retail is not a slow season: one account left in November.p<.01
H2Product F refunds run 3.8× the catalogue average.n=214
✕Weather drives West revenue: tested, discarded.r=.06
HYPOTHESES 41 / SURVIVED 6 / IN THE DECK 3
SO FAR
7,219
DOCUMENTS ANALYSED · EVERY FIGURE COMPUTED TWICE
02 / DELIVERED

Then you get the work.

One upload returns three artefacts. Not a chat window.

Q4 REVIEW · LIVEUPDATED 41s AGO
REVENUE
$54,275
+6.3% QoQ
CUSTOMERS
312
+32 NEW
REFUND RATE
2.4%
+0.9 PTS
MONTHLY REVENUE BY REGION

You dropped something off, someone competent dealt with it, and now you are reviewing the finished work.

Every important number can explain itself.

3 OF 27 FINDINGS SHOWN

Open one and the trail is right there: the file, the rows, the repair, the test.

TRACE
01refunds_export.csv · sheet RMArows 812–1,004
026 SKU spellings mergedPROD-F
03rate 9.1% vs catalogue 2.4%χ² 38.2 · p<.001
✓held after excluding the 2 largest returnsconfirmed
RE-RUN IN CODE BEFORE DISPLAY · 0 / 4 CHECKS FAILED
refunds_export.csvRMA
0812PROD-C48.00
0813PRODF112.50
0814PROD-F89.00
0815PROD-A22.00
0816prod f134.00
0817PROD-F96.50
0818PROD-B31.00
6 SPELLINGS → PROD-F · RATE 9.1%
TRACE
01inventory_snapshot.pdf · pages 4–1162 stockout days
02matched to order gaps in q4_clean.xlsx148 gaps
03baseline: trailing 8-week velocity per SKUper-SKU
✓range $7.4k–$11.2k, midpoint reported$9,100
RE-RUN IN CODE BEFORE DISPLAY · 0 / 4 CHECKS FAILED
inventory_snapshot.pdfP. 4–11
62 STOCKOUT DAYS · 148 ORDER GAPS
TRACE
01survey_q4.xlsx · tenure joined on id1,880 responses
023 rating scales normalised to 1–5repaired
03<6 months 3.4 vs >2 years 4.3t = 6.9
✓onboarding named in free-text answers41%
RE-RUN IN CODE BEFORE DISPLAY · 0 / 4 CHECKS FAILED
survey_q4.xlsxn=1,880
<6 MONTHS3.4
>2 YEARS4.3
ONBOARDING NAMED IN 41% OF FREE TEXT
CAPABILITIES

The whole job, in one place.

Six things that used to be six tools and a Friday afternoon.

Pay for the work,
not another tool.

NO SEATS REQUIRED

Plus

Recommended

$6.67/ month

Billed annually at $79.99 · 33% less than monthly

For the export that lands every month.

$1.33 a project at full use, against $14.99 bought one at a time.

Start Plus
  • 5 projects every month
  • Every export format, Parquet and project JSON included
  • Your logo and colours on every deck, report and dashboard
  • Reports that re-run on a schedule from your saved recipe

Studio

$13.33/ month

Billed annually at $159.99 · 33% less than monthly

For a team, or a book of clients.

$0.67 a project at full use, against $14.99 bought one at a time.

Start Studio
  • 20 projects every month
  • Everything in Plus
  • Four times the projects for twice the price

Your findings open in full for an email (evidence, confidence and limits) before you pay for any of it. Just need one file? Projects start at $4.99.

Questions,
answered.

Start a project

Uploads are deleted 30 days after delivery by default, or immediately on request. Nothing is used for training.

No. Repairing the file is the first stage of the work: duplicates, mixed date formats, money stored as text, totals stranded mid-table. Every repair is listed, and your original file is never modified.

Open its trace: the file, the rows, the repair, the test. Every figure is also computed twice by two independent engines, and anything they disagree about is dropped before you see it.

Not yet. It analyses one table at a time; upload several files and it will tell you what it found in each rather than silently picking one.

No. Leave an email and your findings open in full (evidence, confidence and limits) before you pay for anything.