Who runs it
Karan holds a Bachelor of Engineering in Industrial Engineering from Toronto Metropolitan University and a Lean Six Sigma Black Belt. Industrial engineering is, in practice, the study of how repeatable work should be structured, measured and improved. Recurring finance work, such as combining exports, matching transactions and checking totals, is exactly that kind of work.
How the work is approached
- Understand the process before automating it.Map what actually happens, step by step, including the workarounds. Automating a process nobody has looked at closely tends to automate its problems too.
- Separate the rules from the judgment.Most of a recurring workflow follows rules that can be written down. The rest needs a person. Keeping the two apart is what makes automation safe.
- Build it so it can be tested.Every run checks its own inputs and outputs. Missing or unreadable data stops the run. Differences are shown for investigation and clearly marked as needing review, never presented as a clean result.
- Build it so someone else can maintain it.Documented logic, a change log and a handover, in tools your team already uses wherever practical.
Where the method was tested
The approach was developed during an industry-sponsored engineering capstone at Toronto Metropolitan University. Karan co-led the project with a project partner. It redesigned an enterprise-scale securities reconciliation that an operations team was working through in spreadsheets, using ordered matching rules, rollups and exception workflows on a DMAIC frame. The seven-pass matching engine reached a 99.01% auto-match rate on $1.475B in daily trades, and the redesigned process cut reconciliation cycle time by 87.5%.
Stated precisely
That was an academic capstone with an industry sponsor, not a paid MatchPass deployment, and it was co-led rather than done alone. The 99.01% and 87.5% figures are results from that project, not from a client engagement. Every other figure on this site comes from synthetic demonstrations we built ourselves.
One idea from that project carries into every build: ordering. Run the strictest rule first and let each later pass loosen exactly one condition. A broad rule run early produces matches that look right and are not, and those cost more than an unmatched line, because nobody goes looking for them. Multiple passes, strictest first, is where the name MatchPass comes from.
What this practice deliberately is not
- Not an accounting or assurance service.MatchPass Labs does not provide accounting, tax, audit or assurance services. Accounting judgment and final approval remain with the client.
- Not a software platform.Each engagement delivers a workflow in your environment, documented and handed over so your team can run it without us.
- Not automation for its own sake.Many workflows need nothing more than well-built formulas and Power Query. Code is used where it helps, not as a requirement.
- Not a decision-maker.Matching rules are explicit and deterministic. A language model does not decide whether two transactions reconcile, and people make the judgment calls.
Where we are
MatchPass Labs operates from Toronto, Ontario, Canada as a sole proprietorship, working remotely with organizations in Canada and the United States.