Alteryx Conversion Matrix

PlaidCloud converts Alteryx workflows, apps, and macros into native Advanced workflows. The importer maps each Alteryx tool to a workflow step, macro construct, controlled variable, Document-backed file operation, or managed job executor — so a converted workflow runs cloud-native, not as an emulation.

Nearly every standard Alteryx tool converts fully. The support status below is per tool:

• Full — converts to a native PlaidCloud step, macro, variable, artifact, or managed executor and runs the tool’s function. Most tools are here.
• Partial — converts and runs for its common use, while one distinct capability mode of the tool is not yet reproduced. Rare.
• Not supported — no conversion; the capability is rebuilt natively instead.

External-system connectors (Salesforce, HTTP, Google Analytics, Cassandra, email, Hadoop, Spark, and the like) are connected, not converted — see Connecting to External Systems.

Tool Support

Alteryx Tool	Status	PlaidCloud Equivalent	Converts
Action	Full	Variable binding and conditional step configuration	Updates downstream step settings from converted app inputs.
Analytic App interface (.yxwz)	Partial	Published Panel “App Runner” front-end	The app’s <Questions> layout is materialized into a working Panel app that collects user inputs and drives the converted workflow, published automatically as part of conversion (an app with no input widgets is skipped). Common question types — text, numeric (Numeric Up Down), date, check box, radio button, drop-down, list box, tree, file browse, and folder browse — become App Runner widgets. A question whose value binds to no workflow variable is rendered with a visible note rather than silently collecting a dead value, and less common question types are flagged for review rather than materialized. See Use Converted Alteryx Apps.
Append Fields	Full	Append fields transform	Appends fields from one stream to another.
Auto Field	Full	Auto field sizing transform	Preserves inferred field sizing.
AutoML	Full	ML: Train Model	Fuses the model-method, target, and candidate-algorithm selections into an ML Train step, the same as Classification and Regression. AutoML never records which candidate algorithm actually won at Alteryx run time, so conversion picks the first checked algorithm ML Train supports and notes the choice. Validated against a reconstructed fixture rather than a real Alteryx export — AutoML left Alteryx’s palette after 2023.1, so no export sample exists to convert from; confirm against a real workflow if one surfaces.
Barcode	Full	Barcode executor	Reads or writes barcodes in the configured symbology.
Batch Macro	Partial	Macro fanned out over its control-anchor table	The caller runs the whole macro once per record of its control input, binding the interior’s control parameters each pass — including a control parameter the caller renamed to a display label when exactly one is in play, and an UpdateRawXml Action that rewrites a Select’s field list from a variable. Shapes whose per-pass value cannot be reproduced safely fail closed by name rather than convert wrongly: an UpdateRawXml that rewrites anything other than a Select’s field list, and an Action that replaces a Filter’s entire expression with a variable (its raw per-pass syntax would mis-evaluate).
Browse	Full	Browse / passthrough	Preserved for inspection with no runtime cost.
Buffer	Full	Spatial Buffer	Grows each geometry by a fixed distance.
Build Features	Partial	Build Features	Converts arithmetic and interaction primitives, date-part extraction, explicit-threshold binning, z-score and min-max scaling, and one-hot encoding. Cross-table aggregation (Deep Feature Synthesis across the Manage Relationships tab), PCA, target encoding, and data-driven auto-binning have no equivalent — each is flagged by name in the conversion report rather than silently dropped.
Calgary Cross Count	Full	Calgary database aggregate	Groups indexed fields and counts each custom field’s named values.
Calgary Cross Count Append	Full	Calgary Cross Count Append	Matches each input record against a value index, then counts how many database records it matched.
Calgary Input	Full	Calgary database input	Reads the database with its saved query applied as a filter.
Calgary Join	Full	Calgary Join	Matches each record against a value index on the database, keeping the records that matched.
Calgary Loader	Full	Calgary database writer	Writes the stand-in table every Calgary reader binds to.
Centroid	Full	Centroid executor op	Reduces each geometry to its centre point, appended as a Centroid column.
Check Box	Full	Controlled workflow variable	Converts app check-box choices to controlled input.
Classification	Full	ML: Train Model	Fuses the Assisted Modeling chain into one ML Train step with its algorithm, target, features, and hyperparameters.
Comment / Annotation / Link	Full	Canvas annotation	Preserved as workflow context.
Condition	Full	Step condition	Triggers warnings, errors, or branches from an expression.
Control Parameter	Full	Macro control parameter	Maps to PlaidCloud macro parameters.
Convex Hull	Full	Convex Hull executor op	Builds the smallest convex polygon enclosing each geometry, appended as a ConvexHull column.
Create Points	Full	Table Extract with geom_point	Builds point geometry from longitude/latitude columns, in SQL.
Create Samples	Full	Table Extract, one per output	Splits input into Estimation, Validation, and Holdout at the configured percentages.
Cross Tab	Full	Pivot / cross-tab transform	Pivots rows to columns across every aggregation method and the derived totals.
Data Cleansing	Full	Data cleanse transform	Cleans whitespace, nulls, punctuation, and casing.
Date	Full	Workflow variable (date)	Emits ISO date values for steps and conditions.
DateTime	Full	Date/time transform	Converts date and time parsing and formatting.
Detour	Full	Conditional branch routing	Converts route selection to DAG conditions.
Detour End	Full	Conditional branch merge	Rejoins conditionally selected branches.
Directory	Full	Document directory listing	Lists files from a Document path.
Distance	Full	Table Extract with ST_DISTANCE_SPHERE	Geodesic point-to-point distance and bearing, in SQL.
Download	Full	REST Request step against a REST connection	Downloads external data or artifacts; assign a REST connection to the converted step before it runs.
Drop Down	Full	Controlled workflow variable	Converts app drop-down choices to controlled input.
Dynamic Input	Partial	Dynamic Document input	Resolves file patterns and variable-driven inputs at run time. An Excel ReadList — one worksheet read from every workbook a Directory tool lists, then unioned — converts to a folder-glob sheet read when an upstream Filter pins the sheet to a single name by exact match. A sheet chosen by a Contains, an inequality, or an OR filter, or a read spanning more than one sheet, refuses rather than read the wrong sheet.
Dynamic Rename	Full	Dynamic rename transform	Renames fields from metadata or rules.
Dynamic Replace	Full	Dynamic replace transform	Applies replacement rules from a second stream.
Dynamic Select	Full	Dynamic field selection transform	Selects fields by type, name, or rule.
Error	Full	Step condition (error)	Converts configured error behavior to step conditions.
Feature Types	Full	ML: Train Model feature configuration	Carries each field’s assigned model role into the fused ML Train step.
File Browse	Full	Controlled Document file variable	Lets users choose a file for a converted app run.
Filter	Full	Filter transform	Splits records by expression into true and false paths.
Fit	Full	ML: Train Model	Collapses into the fused ML Train step.
Folder Browse	Full	Controlled Document folder variable	Lets users choose a folder for a converted app run.
Formatted Excel output into a styled template	Partial	Template-aware Excel writer (Export to Excel Sheets for multi-sheet)	Writing into a styled template preserves the template’s formatting — values, number formats, fonts, fills, borders, merged cells, and column widths carry through to the named sheet and range. This covers both a single sheet and several outputs assembled into different sheets of one workbook, filled in a single pass so the sheets no longer overwrite one another. One case stays partial: a template that carries pivot tables or charts drops those objects when its sheets are written. Plain (unstyled) Excel output is Full — see Output Data.
Formula	Full	Formula transform	Converts field expressions to PlaidCloud expressions or SQL.
Fuzzy Match	Full	Fuzzy matching executor	Matches on keys, thresholds, and candidate review.
Generalize	Full	Spatial Generalize	Simplifies geometry to a tolerance, preserving topology.
HTML	Full	Report text / HTML artifact	Preserves content as report or artifact output.
Image Processing	Full	Image transform executor	Applies grayscale, scale, crop, and rotation in canvas order.
Image Profile	Full	Image profile executor	Reports dimensions, mode, format, channels, and luminance statistics.
Image to Text	Full	OCR executor	Extracts text from images through managed OCR.
Input Data	Full	Document-backed file input	Loads .yxdb, .dbf, Excel, and fixed-width .flat source files into workflow data.
Insights	Full	Dashboard / artifact output	Creates a cloud-native review artifact.
Interactive Chart	Full	Chart artifact	Creates a chart artifact from converted data.
Join	Full	Join transform	Produces joined, left-only, and right-only streams on a single- or multi-field key or by position.
Join Multiple	Full	Multi-join transform	Joins multiple input streams.
Key/Value Pairs	Full	Text: NLP AI (keyvalue task)	Extracts key/value pairs from a text column through the LLM-backed NLP AI step, with optional candidate keys and fuzzy or exact matching.
Line To Polygon	Full	Line To Polygon executor op	Closes each line into a polygon ring, appended as a Polygon column.
List Box	Full	Controlled workflow variable	Converts app list selections to controlled input.
Location Optimizer	Partial	Location Optimizer	Converts the linear distance-minimization case — choosing K sites to minimize total or weighted distance to demand — to the native p-median MIP step. A gravity-kernel score, a bespoke scoring macro, or any other candidate-search objective still refuses; see Migrate Alteryx Workflows.
Macro calls	Full	Macro invocation	Imports known macro sources and maps macro calls to PlaidCloud macro steps.
Macro Input / Macro Output	Full	Macro input / output port	Map directly to PlaidCloud macro ports.
Make Grid	Full	Spatial Make Grid	Tiles an extent into square cells, one row per cell.
Map	Full	Map artifact	Creates a PlaidCloud map artifact.
Map Input	Full	Spatial File Import	Reads MapInfo, ESRI, KML, and GeoJSON files with their sidecars.
Master Selector	Partial	Controlled workflow variable and conditional step visibility	Converts the tool’s show/hide visibility conditions on interface elements to conditional step visibility. Converts which sub-workflow the selection runs on a best-effort basis — validation against real on-tenant apps is still pending, so review the selected run-target before relying on it in production.
Message	Full	Step condition (message)	Emits workflow warning, message, or error from a condition.
Modeling	Full	ML: Train Model	Fuses into the ML Train step with the pipeline’s model choice.
Multi-Field Formula	Full	Multi-field formula transform	Applies a formula across selected fields.
Multi-Row Formula	Full	Window / row-aware formula transform	Converts row-relative logic to window behavior, partitioned by Group By.
Named Entity Recognition (NER)	Full	Text: NLP AI (ner task)	Extracts entities (people, organizations, locations, dates, quantities, or a custom type list) from a text column through the LLM-backed NLP AI step.
Numeric Up Down	Full	Controlled numeric variable	Converts app numeric input to a typed variable.
Output Data	Full	Document / table output	Writes output to Document or PlaidCloud tables.
Overlay	Full	Spatial Process	Intersect, union, or cut two geometry columns.
PDF Input	Full	PDF extraction executor	Extracts text or tables from PDFs.
Point To Line	Full	Point To Line executor op	Threads each group’s points into one SequenceLine per group, ordered by the sequence field.
Poly-Build	Full	Spatial Poly-Build	Builds a polygon or convex hull per group of points.
Poly-Split	Full	Spatial Poly-Split	One row per vertex, component polygon, or hole.
Portfolio Composer (Text / Table / Image / Layout / Render)	Full	Report artifacts	Convert report content, layout, and rendering to PlaidCloud report output.
POS Tagger	Full	Text: NLP AI (pos task)	Tags each token’s part of speech from a text column through the LLM-backed NLP AI step.
Predict	Full	ML: Score	Scores data with the trained model table and appends a prediction column.
Radio Button	Full	Controlled workflow variable	Converts app radio choices to controlled input.
Random % Sample	Full	Table Extract (random)	Returns the exact count or percentage of records requested.
Record ID	Full	Row identifier transform	Adds a deterministic record identifier in the configured type and start value, restarting per group.
Redistribute	Full	Redistribute executor op	Reallocates a measure from one set of geographies onto another by area of overlap, appended as a Redistributed column.
RegEx	Full	Regular expression transform	Parses, matches, or replaces text.
Regression	Full	ML: Train Model	Fuses the Assisted Modeling chain into one ML Train step.
Report Map	Full	Map report artifact	Produces a cloud-native map/report artifact.
Run Command	Partial	REST Request step against a REST connection (curl/wget only)	Converts a curl or wget command line to a REST Request step. Interpreter and script commands (cscript, PowerShell, Python, and similar), Excel-COM automation, filesystem operations, and opaque or workflow-launching commands are flagged by name for manual review rather than auto-converted.
Sample	Full	Sample transform	Keeps records by count, percentage, or grouping.
Select	Full	Select and schema projection step	Selects, renames, reorders, and retypes fields.
Smooth	Full	Spatial Smooth	Smooths each geometry over a number of passes.
Sort	Full	Sort transform	Sorts records by configured fields and directions.
Spatial Info	Full	Spatial Info	Area, length, centroid, and bounding rectangle as geodesic measures.
Spatial Match	Full	Spatial Match / Spatial Match (Intersect / Unmatched)	Converts every relationship — Within, Contains, Intersects, Touches, Crosses, Overlaps, and Centroid-In — matched in the database or the workflow engine.
Spatial Process	Full	Spatial Process	Intersect, union, or cut geometry columns.
Summarize	Full	Aggregate transform	Groups and aggregates — sum, count, average, min, max, median, mode, standard deviation, variance, and count distinct.
Tab	Full	App tab grouping	Preserved as converted app structure.
Test	Full	Step condition	Converts test assertions to step conditions.
Text Box	Full	Controlled text variable	Converts app text input to a typed variable.
Text Input	Full	Inline table input	Creates inline data for the workflow.
Text Pre-processing	Full	NLP preprocessing executor	Normalizes and preprocesses text.
Text Summary	Full	Text: NLP AI (summary task)	Summarizes a text column to a configured maximum sentence count through the LLM-backed NLP AI step.
Text To Columns	Full	Split columns transform	Splits text into fields or rows.
Tile	Full	Tile / grouping transform	Assigns tile groups by configured rule.
Tool Container	Full	Canvas container / execution group	Preserved as workflow organization.
Topic Modeling	Full	Topic modeling executor	Runs topic modeling through managed NLP.
Transformation	Full	Transform step	Converts transformation logic to PlaidCloud expressions or SQL.
Transpose	Full	Unpivot transform	Converts columns to rows into a Name/Value pair.
Tree	Full	Controlled workflow variable	Converts app tree selection to controlled input.
Union	Full	Union transform	Combines streams by name, position, or configured rules.
Unique	Full	Unique / duplicate split transform	Separates first-unique records from duplicates.
Visual Layout	Full	Canvas layout metadata	Preserved as design context.
Word Cloud	Full	Text visualization artifact	Creates a visualization artifact from text analysis.
XML Parse	Full	XML parse transform	Extracts XML fields into workflow data.
Zero-shot Text Classification	Full	Text: NLP AI (zeroshot task)	Classifies a text column against a per-row list of candidate labels, single- or multi-label, through the LLM-backed NLP AI step.
Find Nearest	Full	Spatial Find Nearest	Distance-ranked nearest-neighbor join — straight-line, and drive-time nearest (ranked by minutes on a road network) via PlaidCloud’s self-hosted routing engine. Drive-time enabled per workspace.
Trade Area	Full	Spatial Trade Area	Concentric buffers sized in real-world units — fixed-radius, and drive-time trade areas (minutes on a road network) via the self-hosted routing engine. Drive-time enabled per workspace.
Image Template	Full	Image Template executor op	Manual mode crops the image to each region you draw on the template. Automatic mode detects the page regions for you with an open-standard layout model and emits one row per region, each carrying its detected bounds and confidence for review; a page the model cannot read confidently is flagged by name rather than cropped. Open-standard detection can differ from Alteryx’s own — review the detected regions, or use Manual mode for exact control.
Image Recognition	Full	Image recognition executor	Trains an image classifier in the workflow and scores new images against it, in place — no external account and no data leaves PlaidCloud. The classifier is retrained rather than copied from Alteryx, so results are a faithful approximation rather than a bit-for-bit match, and training runs on CPU. Enabled per workspace.
Geocoder	Full	Self-hosted geocoder	Geocodes addresses to coordinates and reverse-geocodes coordinates to addresses against PlaidCloud’s own geocoder — no external account, no data leaves PlaidCloud. Enabled per workspace.

Interface widgets (drop-downs, list boxes, check boxes, text/numeric/date inputs, file and folder pickers), canvas objects (comments, links, containers), and macro ports all convert as native controlled variables or canvas objects. An Alteryx App’s full interface layout — the .yxwz file’s <Questions> definition — is also materialized into a published Panel “App Runner” front-end, so converting the app produces a working input form that collects the same values and drives the converted workflow, not just the individual controlled variables. See the Analytic App interface row above and Use Converted Alteryx Apps.

A note on Excel output. A plain new workbook — data written with default formatting — converts fully. Writing into a styled template, filling named sheets and ranges while preserving their formatting, converts for a single sheet and for several outputs assembled into different sheets of one workbook (the Export to Excel Sheets step fills every sheet in one pass). Templates that carry pivot tables or charts drop those objects when their sheets are written; sheet-level file surgery driven from Run Command (Excel COM/VBS) is flagged by name for manual review rather than auto-converted.

Spatial Tool Coverage

Every tool in the Alteryx Spatial palette has a PlaidCloud route, and every one of those routes is a step you can also build by hand — conversion is a convenience, not the only way in. See Geospatial steps for the full reference and Geospatial Analytics for how they fit together.

Alteryx Spatial Tool	PlaidCloud Route	Runs In
Buffer	Spatial Buffer	Workflow engine
Centroid	Centroid executor op	Workflow engine
Convex Hull	Convex Hull executor op	Workflow engine
Create Points	Table Extract expression	Database
Distance	Table Extract expression	Database
Find Nearest	Spatial Find Nearest	Database; routing service for drive-time
Generalize	Spatial Generalize	Workflow engine
Heat Map	Heat Map (macro)	Workflow engine
Line To Polygon	Line To Polygon executor op	Workflow engine
Make Grid	Spatial Make Grid	Workflow engine
Point To Line	Point To Line executor op	Workflow engine
Poly-Build	Spatial Poly-Build	Workflow engine
Poly-Split	Spatial Poly-Split	Workflow engine
Redistribute	Redistribute executor op	Workflow engine
Smooth	Spatial Smooth	Workflow engine
Spatial Info	Spatial Info	Workflow engine
Spatial Match	Spatial Match	Database
Spatial Match — intersection object, Unmatched	Spatial Match (Intersect / Unmatched)	Workflow engine
Spatial Process	Spatial Process	Workflow engine
Trade Area (fixed-radius + drive-time)	Spatial Trade Area	Workflow engine; routing service for drive-time
Pie Wedge Trade Area	Pie Wedge Trade Area (macro)	Workflow engine
Spatial file input	Spatial File Import	Workflow engine
Summarize — SpatialObjCombine, SpatialObjConvexHull	Spatial Combine	Workflow engine
Report Map	Report Map	Workflow engine

Drive-time trade areas and nearest — Trade Area drive-time rings and Find Nearest drive-time ranking run on PlaidCloud’s self-hosted routing engine (road-network isochrones and travel-time matrices), not a hosted third-party API, so origin coordinates never leave the platform. Drive-time is enabled per workspace. Routing and drive-time areas use OpenStreetMap data — © OpenStreetMap contributors, available under the Open Database License (ODbL).

Known Spatial Prerequisites

The spatial tools themselves convert; these are prerequisites on the input.

• The .geo GeoFile format. Alteryx’s proprietary binary GeoFile has no open reader, so an import of one fails closed rather than crashing at run time. In Alteryx, re-export the layer to a standard spatial format — shapefile or GeoJSON — and import that instead; other spatial formats are unaffected. See Migrate Spatial Alteryx Workflows.
• The Peano key Spatial Info measure. Peano key is a proprietary Alteryx spatial index with no public specification, so a Spatial Info that asks for it skips that one column and names it in a note. Every other Spatial Info measure — area, length, centroid, bounding rectangle, object type, part and point counts, and end points — converts.
• Projected Create Points coordinates. Create Points converts floating-point and integer (×1,000,000) lon/lat coordinates directly. Coordinates already in a projected system need reprojecting to WGS84 lon/lat upstream first, because the workflow does not record which projection they use.

Calgary Tool Coverage

An Alteryx Calgary database (.cydb) is a proprietary indexed store PlaidCloud cannot open. Conversion treats each one as standing for one ordinary PlaidCloud table, named calgary_<database name>, so the tools built around it convert as ordinary table operations rather than refusing outright. See Calgary Databases for how the stand-in table works, and Calgary Join and Cross Count Append for the index-matching tools.

Alteryx Calgary Tool	PlaidCloud Route	Converts
Calgary Loader	Table Extract writing calgary_<database>	Yes, once the database is named and it stores at least one data field.
Calgary Input	Dynamic Document input reading calgary_<database>, with the saved query as filter	Yes, including a query built from an Or or wrapped in a Not, and a contains or starts-with filter (rebuilt as a SQL LIKE). Refuses on a spatial-lookup query and on Skip Records/Max Records limits.
Calgary Input (Count Only) / Calgary Cross Count	Aggregate transform over calgary_<database>	Yes, including a bucket built from an Or or wrapped in a Not. Refuses on a cross over more than one custom field, and on a count-only read naming no column to count over.
Calgary Join	Dynamic Document input matching each record of its input against calgary_<database>	When the incoming field is a plain value matched against a value index — including Count Only mode, which counts the database records each input record matched, a record matching nothing counting zero. Refuses when the field is spatial (rebuild as Spatial Match) or its type is unresolved, and on range-index or unmatched-output-wired Joins, on a Join naming no match field, and on a Join with nothing wired to its input.
Calgary Cross Count Append	Dynamic Document input matching each record, then a counted join over calgary_<database>	When the incoming field is a plain value matched against a value index: it counts, per input record, the database records matched, and a record that matched nothing counts zero. Refuses on a spatial or unresolved index (rebuild as Spatial Match), a custom-value cross-count grid, a range index, and a match naming no field.

Known Calgary Prerequisites

Most of these are prerequisites on the input rather than conversion limits.

• A spatial-index match needs rebuilding as a Spatial Match. A Calgary Join or Cross Count Append matched against a spatial index refuses, naming the stand-in table — the workflow records the index’s name but not whether it holds ordinary values or spatial geometry, and a value match (an equality join) and a spatial match (point-in-polygon) are different operations. Export the database to a table and rebuild the match as a Join (value index) or a Spatial Match (spatial index).
• A read limited by Skip Records or Max Records needs an explicit order. The exported stand-in table carries no record order, so add an order column when you export if you need to take the first or last N records by position.
• A demographic or reference database must be loaded first. A Calgary read before its database has been loaded stops, naming the table to build. The .cydb file itself has no open reader, so re-export it in Alteryx (a Calgary Input into an Output Data tool writing CSV or Parquet) and import that as the named table — see Unreadable .cydb (Calgary) Databases. This is the common case for the licensed demographic and reference .cydb files Alteryx ships (Experian, Census, and similar): no converter can reproduce that third-party data, so bring your own licensed reference data and load it into the named table — the same as the Geocoder reference data.
• A Calgary Loader that stores no data field doesn’t convert, since the table it wrote would have no columns.
• A custom-value cross-count grid doesn’t convert yet. A single plain cross-count field converts — it counts, per input record, the matching database records — but a grid of named custom values would append one count column per value, and Alteryx documents neither how many columns that is nor what it names them. Tracked for a future release, pending a reference export to reverse-engineer the column contract.

Value, range, contains and starts-with queries all convert to a filter; only a spatial-lookup query, an index operation with no column-level equivalent, does not.

Connecting to External Systems

Alteryx’s connector endpoints — Salesforce, HTTP, Google Analytics, Cassandra, email, Hadoop, and Spark, in both directions — do not convert, because PlaidCloud reaches these systems through a connection rather than through a tool on the canvas. Each one refuses by name and states where the work belongs:

Alteryx Tool	Where It Goes in PlaidCloud
Salesforce Input / Output	A Salesforce connection
HTTP Input	A REST Request step against a REST connection
Google Analytics Input	Import: Singer Source with the Google Analytics tap
Cassandra Input	Import: Singer Source with the Cassandra tap
Email Input	Import: Singer Source with the GMail tap
Hadoop Input / Output	Import: SQL / Export: SQL through a Hive, Impala or Presto connection
Spark Input / Output	Import: SQL / Export: SQL through a Databricks connection

Three of these have no write path at all. PlaidCloud reads Google Analytics and Cassandra but never writes to them, so a Google Analytics Output or Cassandra Output tool becomes an Export: SQL to a supported destination instead. Email is not a data destination either: Notify: Email sends a notification to people, so an Email Output tool becomes an Export step plus a separate notification.

Validating a Converted Workflow

For production workflows, validate converted outputs against trusted Alteryx outputs — PlaidCloud validation focuses on schema, row count, and row values, and ignores row order unless the workflow depends on it. Specialized operations (spatial, fuzzy matching, OCR, NLP, reporting) run through managed job executors that keep the workflow cloud-native; machine-learning pipelines convert to the native ML: Train Model and ML: Score steps. See Migrate Alteryx Workflows for the end-to-end import and validation guide.