The task
The group's data was scattered across several dozen sources — websites, exports, different formats and structures. They needed one regularly updated XML feed in a single format that could be handed to downstream systems without manual reassembly.
What we did and what data we collected
We collected data from all the sources — websites, files and exports in various formats — and mapped it to a single schema. In addition, we set up assembly of the final XML feed with validation and scheduled refresh, delivered as DaaS / data via API.
Challenges and how we solved them
The main challenge was heterogeneity. Several dozen sources means dozens of structures: JSON here, tables and CSV there, documents elsewhere, and sometimes only web pages. We described a single target schema and mapped every source onto it.
The second was freshness. Sources update on different cadences, so collection ran on a schedule and changes were picked up automatically.
The third was feed correctness. The final XML was validated against the structure before publication, and duplicates were removed so feed consumers would never receive broken or doubled records.
The result
The group got a single XML feed assembled from dozens of sources automatically and on schedule — instead of manually stitching together scattered exports.
Services in this case study: DaaS and data via API · Web Scraping Services · Website and Data Change Monitoring
Useful reading: XML parsing · JSON parsing · Document parsing · Data normalization