Getting started
Your first project starts with the data you have. Datally handles the preparation and brings the decisions that need you into one place.
Try two or three source files. A reference or target example is optional.
See progress, the decisions made, and any questions waiting for your judgment.
Preview the rows and review any validation findings before using your dataset.
Automatic · the default
You do not need to configure every stage or pick a model to get started.
Step 01
Choose Start a Project from the home screen. Add your CSV or Excel source files. If you already have a target example, add it as the reference. Otherwise, leave the target empty.
Step 02
Keep the default preferences for your first run. Datally analyzes the files, proposes a target when needed, and coordinates matching, translation, validation, and output preparation.
Step 03
If a column match, value, or validation finding needs attention, read the context and accept or correct the proposed choice. Confirm your decisions to continue.
Step 04
Preview the output and check the row counts and findings before downloading CSV or Excel. Datally saves the preparation and creates a reusable Pipeline for subsequent runs.
Advanced workflow
Open Advanced inside a project for the full workbench. Here you can edit and lock individual versions yourself; Automatic coordinates these steps for you.
01
Create a project, add one reference file when you have a target schema, then add every source file you want to reconcile. CSV and Excel sources can use different column names and layouts.
Confirm each intended file appears in the project before continuing.
02
Datally profiles the files and builds the working dictionary. Review names, types, formats, allowed values, and constraints. Make any corrections before locking a dictionary version.
Lock a dictionary version when its target schema is ready for mapping.
03
Review suggested mappings file by file, resolve unmapped columns, and add computed custom columns where a target value must be derived. Keep ambiguous mappings manual.
Lock the mapping only after every required target has an intentional source.
04
Standardize codes, labels, and categories into the dictionary vocabulary. Accept safe suggestions, edit uncertain values, and preserve explicit manual mappings.
Lock the translation version used for the output you intend to create.
05
Author or review validation rules, run them against the selected sources, and inspect exceptions. Correct the source, rule, or accepted exception instead of hiding failed records.
Re-run affected rules after changes and review the current issue totals.
06
Select the intended source files and locked versions, choose exact-row duplicate handling, then consolidate. Resolve any surfaced conflicts and download the accepted CSV or Excel output.
Verify row counts, source coverage, mapping coverage, and validation evidence before use.
Optional local AI
Automatic selects compatible available models for uncertain cases. You can inspect its choices and use Advanced settings when you want to choose explicitly.
Install Ollama and download a model suited to your hardware to keep AI inference on your machine. If no compatible model is available, preparation still runs with deterministic matching and review.
Local AI setupTroubleshooting
Open the review queue, resolve the pending decisions, then continue. The default review policy also asks for your approval when checks find errors.
Reopen the project to see its saved progress and decisions. If the attempt was interrupted, use Resume or Try again when offered.
Automatic continues with deterministic matching and asks you to review unresolved choices. To add local AI, follow the optional setup guide below. Cloud AI must be enabled deliberately.