Engagements
Priced from a floor, scoped in writing.
Three analytics engagements and one AI add-on. Scope is agreed in writing before work starts, and the deliverables below are what you actually receive — not a category of activity.
| Engagement | Duration | Main deliverable | From |
|---|---|---|---|
| Reporting audit | Two weeks | An inventory of every system behind your board, risk and client reporting, and who owns each one | Placeholder · Price |
| Data platform build | Six to ten weeks | Automated ingestion from custodians, fund administrators and internal systems, scheduled and logged | Placeholder · Price |
| Board & risk reporting | Four to six weeks | Written definitions for every figure, agreed with finance, risk and the front office | Placeholder · Price |
| AI assistant on your dataAdd-on | Two to three weeks | A chat assistant connected only to your approved reports and definitions | Placeholder · Price |
01
Reporting audit
From Placeholder · Pricefixed fee · Two weeks
Two weeks to find which of your numbers disagree, why they disagree, and what it takes to make them agree for good.
Why start here
Most reporting disputes in a bank or an asset manager are not tooling problems. They are definition problems — two teams using the same word for different things — or ownership problems, or a spreadsheet between the custodian file and the board pack that nobody dares to touch. Until you know which one you have, buying software is guessing.
The audit is deliberately short and fixed in price. Two weeks, a fixed fee, and a document at the end that you own outright, whether or not we work together again.
How it runs
I start from the report your committee actually reads and work backwards. Every figure on it is traced to the system it came from. Where two systems disagree, I find out why, write down which one is right, and note what it would take to make that permanent.
You get an hour of my time each week for questions, and a written summary at the end. No slideware.
What it does not include
The audit fixes nothing. It tells you precisely what is broken and what fixing it costs. Some firms take that document to their own data team, and that is a perfectly good outcome: the document is written so it is possible.
02
Data platform build
From Placeholder · Pricefixed scope · Six to ten weeks
One validated source for positions, transactions and client data, with checks that stop bad numbers before they reach a report.
What gets built
A report is only as honest as the data platform under it. This is that platform: the scheduled, tested, unglamorous machinery that takes whatever your custodians, administrators and internal systems send, and produces one dataset that finance, risk and the front office all accept.
How breaks are handled
A platform that quietly serves stale or unreconciled numbers is worse than none, because people sign off on it. Every build asserts on the data’s own timestamps and stops rather than publishing a stale snapshot. Every reconciliation break is recorded with its cause and the person who owns it. If a feed dies overnight, you find out in the morning from an alert, not at quarter end from an auditor.
What you own at the end
Everything is written in Python, with tests, in your repository. No proprietary runtime, no per-seat licence, nothing that stops working when the engagement ends. The handover session is part of the scope, not an extra.
If your team has no one to hand it to, say so early. That changes the design, and it is better to know at the start than at the end.
03
Board & risk reporting
From Placeholder · Pricefixed scope · Four to six weeks
Dashboards and reporting packs your board and risk committee read without arguing about the inputs.
Definitions first
The failure this engagement fixes is not a missing chart. It is a committee meeting that spends its first twenty minutes establishing whose exposure figure is correct.
So the first step is not a dashboard. It is writing down what each figure means — exposure, assets under management, net flows — and getting finance, risk and the front office to agree on that sentence in writing.
The pack
One pack, built on data you already trust, designed around the decisions your board and committees actually make. Not a wall of every metric that could be plotted: the handful that change what somebody does next.
Colour carries meaning or it stays out. Every palette is checked for colour-blind separation, and every chart has a text alternative, so the pack survives being printed, read aloud or opened with a screen reader.
Keeping it right
Every figure links back to its definition and its source. When a number moves, the pack shows why, so nobody has to go hunting before the meeting. The method note is written for the people who will question it: internal audit, external auditors and, where relevant, your regulator.
+ Add-on
AI assistant on your data
From Placeholder · Priceadd-on · Two to three weeks
A simple, governed chat assistant that answers questions from your own reports and definitions.
What it is
A question-and-answer layer on top of reporting you already trust. People ask “why did net flows fall last month?” in plain language, and get an answer drawn from your own reports and written definitions, with a link to the source. It works best after an audit or a platform build, when the numbers underneath are settled.
What it is not
It does not make decisions, move money, send messages or take any action by itself. It does not see data the person asking could not already see. It is not a replacement for the report: it is a faster way to find the right page of it.
How it is kept safe
Every answer cites where it came from. Outside its approved documents, it says it does not know instead of guessing. Conversations can be logged for your compliance team to review, and the whole set-up is documented before it goes live.
Not sure which one?
Start with the call. If the audit is not the right first step, I will say so — including when the honest answer is that your team can fix this without hiring anyone.