The problem

Monthly HST returns across a group of small holding companies were prepared by hand: export the numbers from the ERP, key them into a return template, then key the same numbers a second time into the CRA portal. Every re-key was another chance to transpose a digit.

What I built

A macro that ingests the month's financial information from the ERP and populates the HST return template in Excel. After a human review, a second tool populates the CRA web form directly. The filer logs in with the BN, reviews what's on screen and clicks submit — control stays entirely with the person. The tool only moves the data.

The problem

A transaction due diligence required every employee form previously uploaded to the client's HRIS. The software vendor was approached for a bulk export and quoted north of $300,000 to produce it.

What I built

An automated tool that works through the HRIS employee by employee, downloads every form attached to each record, and saves them into a consistent folder structure ready for the data room.

The problem

A client was spending manual time each period preparing compliance notices to banks across multiple debt certificates. The underlying financial data already lived in an Excel tracker — it was the reading, transcribing and formatting that consumed the hours.

What I built

A pre-programmed AI skill that analyses the Excel financial tracker, pulls out the relevant details for each certificate, and populates the compliance certificate in a consistent format every single time.

The problem

Reviewing external valuation reports meant reading the current and prior year side by side to work out what actually moved, then separately checking the appraisal against recent comparable transactions — and hoping none had been missed.

What I built

A tool that ingests the current and prior year valuation reports and produces a consistent AI-generated summary of the key year-over-year changes. It also takes in a listing of known transactions, identifies whether the appraisal sits inside or outside the range of recent comparables, and flags any that are missing.

The problem

Month-end consolidation across entities sitting on different GL systems meant exporting each trial balance by hand, pasting it into a master workbook, and re-mapping the chart of accounts every period before eliminations could even start.

What I built

A pipeline that auto-ingests trial balances from each GL system on a month-end schedule, normalises them to a single chart of accounts, applies inter-entity eliminations by rule, and produces the consolidated financial report at the end.

The problem

The same routine month-end entries — service charges, wire fees, interest income — were found by scrolling through bank statements and typed into the ERP one line at a time, every single month.

What I built

An automated pipeline that downloads the bank statements, AI-classifies the routine entries, and generates an ERP-ready import template complete with GL codes and cost-centre allocations.