Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a custom python script to match your two data sources
Mazharul Islam
About this gig
Two files. Different systems. Same underlying records but nothing
lines up.
I build a documented, reusable Python script that matches your own two
data sources: deterministic matching first (exact key, like an ID or
email), fuzzy fallback where formatting differs, and duplicate
detection before anything is counted twice.
Not limited to e-commerce the same logic works for CRM-vs-mailing-list
cleanup or vendor-invoice-vs-PO reconciliation.
20+ years of SQL and data-reconciliation discipline, applied to Python.
Every match is checked, not guessed the script flags what it can't
confidently match instead of pairing records on weak evidence.
Included: a documented, reusable script matching two files you provide,
with exact + fuzzy matching and duplicate/exception flagging.
Not included: live database or account access, hosted infrastructure,
or automated pipelines pulling from live APIs.
Send a sample of your two data sources and I'll tell you honestly
whether this fits before you order.
Get to know Mazharul Islam
Mazharul Islam
Shopify and Amazon Reconciliation, Accurate SQL Queries, Views and Procedures
- FromBangladesh
- Member sinceJun 2026
Languages
Bengali, English
Founder of QueryForge Studio. I reconcile Shopify, Etsy, Amazon, and
WooCommerce sales data into clean, profit-by-product reports — catching
price and fee mismatches most sellers never check for.
I also write SQL queries, views, stored procedures, and functions for
businesses that need reliable, accurate database logic — not just fast
code.
20+ years of hands-on SQL experience in banking payment systems (cheque
clearing, EFT, RTGS), on the development side.
Fast, honest turnaround. No unnecessary complexity.
FAQ
1. Do I need to give you live access to my systems?
No. You share two file exports (CSV/Excel) — never live database or account access. I build and test the script against those, you can rerun it yourself.
2. How do I know the fuzzy matches are actually correct?
Every fuzzy match includes the confidence basis (e.g. name similarity score) in the output, and nothing is paired without meeting a real threshold. Anything uncertain is flagged as unmatched, not guessed.
3. Can this handle data that's not from e-commerce platforms?
Yes — it's built to match any two of your own data sources. Past examples include a CRM-vs-mailing-list cleanup and a vendor-invoice-vs-purchase-order reconciliation.
4. Do you host or run this on an ongoing basis?
No. You receive the script and documentation — you run it yourself, whenever you need to. Nothing is hosted or automated on our end after delivery.
5. What if my two data sources don't have a clean shared field to match on?
Message me first with a description of both files. If there's no reasonable basis for matching, I'll tell you that honestly rather than build something unreliable.
