Infosys Data Workbench vs OpenRefine

Infosys Data Workbench (iDW) combines large-scale data quality operations with lightweight analytical master data management. Its AI-powered core supports data-quality processing, while an Apache Spark core supports high-speed analysis at scale. Users can profile data with machine-learning models, review interactive reports, and validate complex business rules. Cleansing and standardization tools include machine-learning-based enrichment, outlier detection, clustering, and filtering. Master-data functions cover de-duplication, linking, grouping, golden-record generation, and identity resolution.

The product works with JDBC-compliant databases and files, including fixed-width, delimited, and Excel formats. Deployment options include standalone or cluster execution, a web interface, and cloud-ready installation; the listed environments include Windows, Linux/Unix, Azure, and AWS. Jobs can run once or recur, with results in dashboards or downloadable PDF and Excel reports. Infosys lists LDAP authentication, single sign-on, and role-based authorization. Pricing is on request, and Infosys directs interested organizations to its request-for-services form.

OpenRefine is a free, open-source tool for cleaning, transforming, and extending messy data. It works on the user's machine rather than sending data to a cloud service, and basic functions can run offline through a small web server on that computer. Facets help explore large datasets and apply changes to filtered views. Clustering uses heuristics to find similar values and address inconsistencies; reconciliation matches values to external databases through reconciliation services. Users can undo and redo operations without a stated history limit. OpenRefine imports formats including CSV, TSV, fixed-width text, JSON, XML, spreadsheets, PC-Axis, MARC, RDF formats, and Wikitext. It can also upload data to Google Sheets and create or update Wikidata statements. Releases are available for Windows, Mac, and Linux. Java must be installed and configured, although the Mac version and some Windows packages include it. Data is not encrypted by default; operating-system full-disk encryption can protect it. The project offers support through its forum, GitHub, community calls, and Gitter chat.