Operations automation for wealth management teams.
Small, precise tools that keep data reconciled across your systems, so your team spends less time checking numbers and more time acting on them.
Small, precise tools that keep data reconciled across your systems, so your team spends less time checking numbers and more time acting on them.
Narrow tools for the checks your team already does by hand. Each one does a single job, runs on a schedule, and tells you when something doesn't agree.
Compare positions, asset mix, cash, or fee balances across two systems and flag every difference, to the basis point, with the likely cause where it can be inferred.
Custodian file against portfolio system. Portfolio system against the general ledger. Model against actual.
Reports that assemble themselves from your existing exports on a schedule, instead of being rebuilt by hand each period.
Month-end packages, mandate drift, cash forecasts, client-ready summaries.
Recalculate fees against the agreed schedule and tiers, confirm what was actually billed, and surface every account where the two don't match.
Tiered and household-aggregated schedules, minimums, and one-off adjustments included.
Standardize naming, remove duplicates, validate fields against your own rules, and leave a log of exactly what changed and why.
Client master files, security identifiers, account mappings between systems.
Take a team that keeps asset-mix percentages in two places: the custodian's holdings file and their own portfolio system.
Every week someone exports both, lines them up in a spreadsheet, and reads down the columns looking for sleeves that don't agree. Most weeks everything matches. The weeks it doesn't, the difference is found late, after a report has gone out or a rebalance has been sized on the wrong number.
The replacement is a script. It pulls both files, normalizes the sleeve names, compares every percentage to the basis point, and writes a short exceptions list: which account, which sleeve, how large the difference, and the most likely cause. It runs before anyone arrives. The person who used to do the comparison now spends that time on the two or three lines that actually need a decision.
Four steps, in this order, every time. You see working output on your own kind of data before you commit, and the work is priced as a fixed cost before it starts.
A short call. You describe the check or report your team does by hand; we agree on exactly what the tool should produce, where the data comes from, and what it must never touch.
We build a demo against anonymized or synthetic data shaped like your files. You see real output, on your own terms, before agreeing to anything further.
The tool runs in parallel with the manual process until the two agree and the people who own the numbers trust it. Differences get investigated, not explained away.
You get the code, a plain-language guide, and a runbook for whoever owns it next. Nothing depends on us staying involved.
Scope and price are fixed after step one. If the demo doesn't do what we agreed, you don't proceed.
Lavallee & Co. is Benjamin Lavallee and a small bench of collaborators from Waterloo.
Ben is completing the Computer Science and Business Administration double degree in Waterloo and spent a co-op term in investment operations at a wealth management firm, doing the reconciliations and reports these tools now do. We know what a custodian file looks like at 7 a.m. and which column is usually wrong.
We work part-time through the school term, from Toronto and Waterloo. That is why the work is deliberately small and scoped: one tool, one job, done properly and handed over. Ben leads every project and stays accountable for the result; when one needs more hands, the bench steps in. You can reach him directly at ben@lavalleeco.ca.
Describe the check or report your team does by hand. A sentence is enough. We'll reply with whether it's a fit and what a demo would take.