Struktive FAQ — Frequently Asked Questions
Frequently Asked Questions
What does Struktive actually do?
Struktive takes a messy CSV export of your asset inventory and returns a clean, structured, scored version of that data — ready to import into any DCIM, ITSM, or CMDB platform. It normalises vendor names (resolving 40–80 raw variants down to 15–25 canonical names), parses free-text location strings into a structured hierarchy (Site → Building → Floor → Room → Row → Rack → U Position), classifies every asset into one of eight categories (Compute, Networking, Storage, Power, Cooling, Facilities, Out of Scope, Unknown), and scores each record 0–100 on data quality. The output is a six-sheet Compliance Audit Pack plus platform-specific export files for NetBox, dcTrack, Device42, or Struktive Native format.
What file formats does Struktive accept?
Struktive accepts CSV and Excel files. It handles real-world exports — files with metadata preamble rows, inconsistent quoting, mixed delimiters, short rows, and empty rows are all handled gracefully. The pre-processor strips noise before normalisation begins. Maximum file size is 50 MB per upload.
Do I need to format my CSV before uploading?
No. Struktive auto-detects column mappings using a library of 200+ header aliases (e.g. 'Cab Location', 'Cabinet', 'Rack', 'Asset Tag', 'Serial Number', 'S/N'). You do not need to rename columns or reformat your file before uploading. If a column cannot be mapped, it is passed through as-is.
How does the quality score work?
Each record is scored 0–100 based on seven factors: field completeness, vendor resolution confidence, location parse depth, classification confidence, serial number validity, duplicate status, and data consistency. Records scoring below 70 are flagged in the Exceptions sheet of the Compliance Audit Pack.
What is the Compliance Audit Pack?
The Compliance Audit Pack is a six-sheet XLSX workbook: Cover (chain-of-custody metadata), Asset Register (normalised records), Audit Trail (field-level transformation log), Duplicates Register, Exceptions Log (grouped by issue type), and Summary Statistics. It maps to SOC 2, ISO 27001, EU EED, NIS2, and TIA-942 requirements.
Does Struktive use my data to train models?
No. Your uploaded data is used solely to produce your normalisation output. It is not used to train any model, shared with third parties, or retained beyond the stated retention period (30 days for job data, 7 days for uploaded files).
How does the NetBox export work?
The NetBox export generates a slug-based CSV validated against the NetBox DeviceType Library. Matched devices get the correct manufacturer slug and device type slug. Unmatched devices are flagged in a pre-flight report that shows exactly which manufacturers, device types, sites, and racks need to be created in your NetBox instance before running the import.
How does Struktive pricing work?
Struktive charges per record. The first 350 records are free with no account required. Paid tiers start at $0.05 per record for up to 2,000 records, with volume discounts for larger datasets. There is no subscription — you pay for each normalisation job.
What counts as a record?
One record equals one row in your uploaded file, after the pre-processor strips empty rows and metadata preamble rows. Duplicate rows are counted once.