Repeatable cleaning pipelines
Power Query or Python transformations you can rerun, not a one-time manual cleanup.
United States · New York City · Data cleaning
The recurring New York data problem is entity matching: the same portfolio company, counterparty or property appearing under four spellings across a custodian file, an admin statement and an internal tracker. Totals agree, detail does not, and the difference only shows up in a drill-down.
Abhishek
Abhishek Singh
Akhilesh
Prahalad
ShaktiA remote team of Excel and spreadsheet specialists. You work directly with the person building your file, from the first question to the final handover. More about us.
A typical job here
Typically 8–16 hours. Unusual cases land outside that, and we say so in the estimate rather than discovering it halfway through.
| Reconcile a custodian file against an internal position tracker | 8–16 hours |
|---|
Why New York City specifically
Clients for this are typically in Midtown, the Financial District, Long Island City and Downtown Brooklyn, across investment and asset management, commercial real estate, media, and professional services firms billing by the hour. Deal models, portfolio packs and month-end consolidations that have to survive review.
For the wider picture of what we see in this market — the recurring problems, the source systems, how the time difference plays out — the New York City guide covers it properly rather than repeating it on every service page.
8–16 hours is typical here. You get a number in writing before we start.
Unlocked files, documentation, recorded handover.
Full guide
This page covers one service. The city guide covers the whole picture, including the problems we see most in this market.
Excel experts in New York CityScope
Two of the things this covers most often for New York City clients. The data cleaning page sets out the full scope and method.
Power Query or Python transformations you can rerun, not a one-time manual cleanup.
Fuzzy matching across sources that never quite agree on how a name is spelled.
What counts as a duplicate, what happens to a blank, how a bad date is handled — written down and agreed.
The records that could not be resolved automatically, listed rather than quietly dropped.
Need to check a feature before hiring? Read Microsoft’s official Excel help and learning for formulas, PivotTables, importing data and troubleshooting.
Questions
Answers for this service in this market, including the ones that might cost us the job.
Ask us something elseUsually yes, because a small recurring difference is rarely random — it is a rule difference that will eventually produce a large one. The useful deliverable is the explanation of why they differ, not a merged file that hides it.
That is the point. Power Query steps are visible and rerunnable, and we document the process. If you would rather we run it each cycle we can, but you should not need us to.
We keep assumptions in one place, avoid hard-coded numbers inside formulas, add reconciliation checks that fail loudly, and write a short update procedure that lives in the file. None of that is exotic — it is just discipline that gets skipped when a model grows under deadline pressure.
Related
Most projects touch more than one of these. We scope them together rather than selling them separately.
Same service elsewhere
Start with the problem
Send the workbook or describe the problem. An hours estimate back within a day, at no cost and no obligation.
Post your projectA short introduction
A short introduction to our Excel and spreadsheet work.
The short introductory video shows spreadsheet tables and dashboards, a person working on a laptop, and a message inviting viewers to post their project.