Who runs it
One person, and the reasons that is deliberate.
I am Ibrahim Shaikh. My background is credit analysis and portfolio risk. Hearth Advisory is the practice I run on my own, and CreditIntel is the offer it leads with.
The person
Credit work first, tooling second.
Most of my working life has gone into one narrow problem. A finance dataset arrives in whatever shape some system decided to hand over, and somebody has to get three things out of it: who owes money, whether they are settling on time, and what has shifted since anyone last looked. NetSuite, Snowflake, SQL and Python are where I spend my time, mainly because that is where the answer tends to be hiding.
Which half I learned first is worth stating, because it shows up in the output. The credit reasoning came before the tooling, not after it. So when I am deciding what belongs in a review, I am working backwards from arguments I have actually sat through in a meeting, rather than forwards from whatever happens to be convenient to chart.
Credit is where I start with a client for the same reason. It is the ground I can stand on fastest and defend line by line. The way of working underneath it is not specific to receivables at all, which is what the Finance Workflow Sprint is there for.
The practice
Owner led, which is a limit as much as a promise.
The person who scopes the engagement is the person who writes the queries and the person whose name goes on the document. Your file is not passed to someone else after the first call. There is no team to pass it to, and I have not built one on purpose.
The honest consequence of that is capacity. One person working carefully can hold a small number of engagements at a time, which is why everything here is sold as a bounded piece of work with a defined end rather than as a standing monthly arrangement. A rolling retainer would suit my cash flow considerably better than it would suit the work, so it is not on offer.
What you are hiring
- direct contact with whoever is doing the analysis, at every stage
- a reconciliation step with the authority to stop a delivery
- a layout held still between runs, so two reviews can be compared
- a named human accepting responsibility for each judgment call
- a boundary agreed in writing before anything starts
What you are not hiring
- an agency with an account layer between you and the work
- a platform you have to adopt or maintain
- a system that makes credit decisions on its own
- an open ended monthly retainer
- a promise about an outcome I do not control
Evidence
Invented examples, honestly labeled.
There is no real company anywhere in the demonstration portfolio, because I decided early that inventing a customer was more honest than implying one. Every name, balance, rating and date in it was generated. It is then checked back against its own generated source by script, and a build that disagrees with that source does not complete, which is the identical rule paid work runs under.
The same decision explains what is absent from this site. No logos, no quotes attributed to clients, no star ratings, no figures attached to an engagement. If a real result ever arrives that I have permission to describe, it will carry a date and a plain note about what it does not prove.
The one piece of live evidence Hearth is building in public is the Local Search Clarity run on this site, where the baseline, the changes and the dated measurements are published as they happen, including the measurements that show no movement.
How I work
Three habits you would notice working with me.
I start at the extract, not the layout
Designing the output is the enjoyable part, so it is the part I have taught myself to postpone. The first block of time goes on writing down what each column means and confirming that with whoever owns the source. A document that reads calmly is a downstream effect of getting that right, and it cannot be bolted on later.
I would rather point at four accounts than four hundred
Narrowing is the actual work. Almost anything can be charted, which makes charting everything a comfortable way of avoiding the decision about what deserves attention this week. I would rather hand over a short list with reasons attached than a display that covers the whole portfolio and commits to nothing.
I mark my own guesses
Where I have concluded something the data does not strictly prove, it is labeled in the document as inference. That makes the deliverable slightly less impressive to read and a good deal safer to act on, and it keeps the decision with the person whose signature is going on it.
Start here
If it is not a fit, I will say so.
The fastest way to find out is a short description of the portfolio or the workflow. One email, one honest answer, no sequence afterwards.