We build inspectable data pipelines, precision spreadsheet audits, and structured operational workflows for growing businesses, bookkeepers, and operators who cannot afford silent data corruption.
All spreadsheet cleaning packages include non-destructive duplicate detection, cell-level change logs, exceptions reports, and mathematical reconciliation. Choose your tier below to order directly via Fiverr escrow protection.
Ideal for small tabular files needing whitespace normalization, duplicate flagging, and date standardization.
Comprehensive data normalization, duplicate detection, and conflict isolation for operational datasets.
Full-scale standardization, deduplication, and schema validation for customer, inventory, or order catalogs.
Operating in the Ottawa National Capital Region? We offer a local starter package for small business operators, bookkeepers, and trades: CA$29 fixed for 1 UTF-8 CSV (up to 100 rows, 6 columns, 1 MiB) with 2 agreed cleanup rules and an exceptions review sheet.
Instant purchase via secure Stripe-hosted payment pages. Scope confirmation and secure transfer instructions sent by email within one business day post-checkout (direct order terms; not platform escrow).
One CSV file up to 50 data rows. A cleaned copy, change log with every original value, and exceptions report.
Stripe Checkout (CA$15) →One CSV file up to 2,000 data rows. Read-only report of duplicate groups, row numbers, and input SHA-256 fingerprint.
Stripe Checkout (CA$29) →Ranked audit and column profile (up to 2,000 rows; 100 rows/30 cols/1 MiB). Read-only data health findings.
Stripe Checkout (CA$39) →Automatic duplicate removal can delete entire rows before conflicting values have been reviewed. We take a non-destructive approach: records are preserved, formatting changes are logged, and duplicates or conflicting values are isolated in an exceptions report for your final decision.
Transform messy exports into structured, uniform datasets. We normalize padding whitespace, align inconsistent dates into standard ISO formats (YYYY-MM-DD), and standardize numeric delimiters with a reversible cell-by-cell modification trail.
Prepare reliable tabular records for internal databases, ERPs, or bookkeeping workflows. We structure multi-source CSV files, enforce required headers, preserve leading zeros, and neutralize CSV formula injection hazards.
We do not treat your data as a black box. Our tools are engineered to be reproducible, mathematically verifiable, and completely inspectable at every transformation step.
Our audit and standardization workflows run on single-purpose, tested Python and Node pipelines that execute deterministically without opaque hallucinations or unlogged modifications.
Every file is validated against an exact row-in equals row-out rule. We mathematically reconcile total input records against total output records so zero rows are lost.
We neutralize dangerous CSV injection vectors by safely escaping unquoted leading symbols (such as =, +, -, @, and |) that can trigger remote code execution when opened in Excel.
Client spreadsheets are processed in an isolated local environment—never uploaded to public LLMs or cloud training sets—and permanently wiped post-delivery with a signed deletion receipt.
On every project, you receive four distinct, checkable files:
cleaned.csv)
Your original tabular data, standardized and preserved with zero rows deleted.
changes.csv)
Every adjustment recorded cell-by-cell showing row, column, rule, and original value.
exceptions.csv)
Duplicates, contradictory records, and ambiguous fields isolated for your review.
summary.md)
Mathematical reconciliation proving input rows match output rows and documenting file hashes.
Inspect our exact synthetic demonstration data. See how conventional deduplication deletes critical records, and how our audit process preserves your data without erasing a single row.
| Row | SKU | Description | Quantity | Unit Price | Date | Flaw Present |
|---|---|---|---|---|---|---|
| 2 | SKU-001 |
" Blue widget " |
10 | "1,250.00" |
2026/10/23 | Extra padding spaces; quoted formatted number |
| 3 | SKU-001 |
" Blue widget " |
10 | "1,250.00" |
2026/10/23 | Exact duplicate of Row 2 |
| 4 | SKU-001 |
Blue widget |
11 | 1250.00 |
2026-10-23 | Conflicting quantity (11 vs 10) on same SKU |
| 5 | SKU-001 |
Blue widget |
10 | 1250.00 |
2026-10-23 | ISO date formatting |
| 6 | SKU-002 |
Cable |
5 | 20.00 |
2026-10-23 | Standard entry |
| 7 | SKU-002 |
Cable |
5 | 20.00 |
2026-10-23 | Exact duplicate of Row 6 |
| 8 | SKU-003 |
Bracket |
0 | 0.00 |
2026-10-24 | Zero quantity entry |
| 9 | SKU-004 |
(empty) | 3 | ? |
10/11/2026 | Missing description; unparseable price symbol (?) |
Raw synthetic inventory dataset showing 8 data records with whitespace, duplicate, and ambiguity defects.
| Row | SKU | Description | Quantity | Unit Price | Date | Preservation Status |
|---|---|---|---|---|---|---|
| 2 | SKU-001 |
Blue widget |
10 | 1,250.00 | 2026/10/23 | Whitespace normalized; preserved |
| 3 | SKU-001 |
Blue widget |
10 | 1,250.00 | 2026/10/23 | Preserved (0 rows deleted) |
| 4 | SKU-001 |
Blue widget |
11 | 1250.00 | 2026-10-23 | Preserved; conflict isolated in audit |
| 5 | SKU-001 |
Blue widget |
10 | 1250.00 | 2026-10-23 | Preserved |
| 6 | SKU-002 |
Cable |
5 | 20.00 | 2026-10-23 | Preserved |
| 7 | SKU-002 |
Cable |
5 | 20.00 | 2026-10-23 | Preserved (0 rows deleted) |
| 8 | SKU-003 |
Bracket |
0 | 0.00 | 2026-10-24 | Preserved |
| 9 | SKU-004 |
(empty) | 3 | ? |
10/11/2026 | Preserved; ambiguous fields flagged for human review |
Cleaned master table: 100% of rows preserved, whitespace trimmed, and zero records silently dropped.
| Row | Column | Transformation Rule | Original Value | Standardized Value |
|---|---|---|---|---|
| 2 | description |
whitespace_trim |
" Blue widget " |
"Blue widget" |
| 3 | description |
whitespace_trim |
" Blue widget " |
"Blue widget" |
Cell-by-cell modification trail documenting every single normalization so changes can be audited or reversed.
| Row | Target Column | Exception Category | Audit Detail |
|---|---|---|---|
| 3 | (whole row) |
duplicate-row |
Identical to Row 2; both rows preserved, flagged for client confirmation. |
| 7 | (whole row) |
duplicate-row |
Identical to Row 6; both rows preserved, flagged for client confirmation. |
| 9 | description |
missing-value |
Empty cell; isolated for human review rather than guessing data. |
| 2, 3 | unit_price |
ambiguous-value |
Formatted currency string with comma; preserved without lossy type coercion. |
| 9 | unit_price |
ambiguous-value |
Contains unparseable symbol ?; isolated rather than coerced to zero. |
Exceptions Review Sheet isolating ambiguities, conflicts, and duplicates for explicit owner sign-off.
Reconciliation Verified: Input row count matches output row count exactly. Zero data loss.
Tools like Excel "Remove Duplicates" make silent, irreversible assumptions:
A deterministic, non-destructive process designed to protect business records:
Ground-level operational knowledge for operators, small businesses, and ecommerce teams managing mission-critical spreadsheets.
How naive "first-row-wins" deduplication silently deletes updated shipping addresses, phone numbers, and recent orders without an audit trail.
A deep dive into leading zero loss, transposed calendar dates, unescaped multi-line text, encoding mismatches, and scientific notation errors.
Five critical pre-import checks for Shopify merchants to prevent catalog splitting, inventory overwrites, and decimal pricing bugs.
Data Integrity & Operational Systems
📍 Ottawa, Ontario, Canada
Operating from Ottawa, Ontario, Dustin Rebellato provides data integrity auditing, spreadsheet standardization, and structured operational systems for growing businesses, bookkeepers, and operators.
Dustin’s background bridges five years of demanding physical operations in construction, demolition, and document control with competitive athletic coaching and computational data architecture. In physical construction, cutting corners or relying on unverified assumptions leads to catastrophic structural failures. The same discipline applies directly to data operations: a single corrupted delimiter, dropped leading zero, or silently deleted customer record can cripple a business's operations.
“Less waste. More intelligence. Greater human benefit. We do not invent complexity to justify software fees. We build lean, verifiable systems that solve real operational problems and keep humans in complete control of their decisions.”
Every client engagement is governed by strict technical ethics: we do not accept confidential data without clear scope boundaries, we process files locally under strict air-gapped security, and we ensure every single deliverable is checkable down to the individual cell.
Tell us about your data or operational challenge. We review your requirements and confirm scope, compatibility, and turnaround within 24 hours before any order is accepted.
This client-side scope tool compiles your non-confidential project requirements into a structured inquiry package. It does NOT upload files or data to unverified servers. Review your summary below, then send it to our operational inbox or use our verified platform channels.
Review your structured project summary. You can send it directly to our operations inbox or copy it below:
Primary communications are reviewed directly by Dustin Rebellato:
obsidian1change1the1world@gmail.com
Choose your preferred route: order via Fiverr's protected milestone escrow, or connect on Kijiji for Ottawa local service: