The Real Cost of a Missed Insurance Renewal

Most families don't think about document organisation until something goes wrong. Not in a dramatic way -- nobody wakes up thinking about their home insurance policy. The cost shows up quietly: a renewal you forgot, a warranty you couldn't find, a tax deduction you didn't claim because the receipt was somewhere in a shoebox.

We built Archevi because we kept seeing the same pattern in our own lives. The documents exist. The information is there. But finding it at the right moment -- that's where families lose time and money.

The Renewal Problem

Here's a scenario that plays out in Canadian households every month. Your auto insurance renewal arrives in the mail. The premium went up $47/month, but you're busy, so you set it aside. By the time you remember, the grace period has passed. Now you're shopping for new coverage with a gap in your history, and every quote is higher than what you were paying.

The Insurance Bureau of Canada estimates that a 30-day lapse in auto insurance can increase premiums by 15-25% for the next three years. On a $2,400/year policy, that's $360-600 in extra costs -- because a piece of paper sat on the counter too long.

This isn't about being careless. It's about the volume of documents a family manages. Insurance policies, OHIP cards, provincial driver's licences, property tax assessments, RESP contribution receipts, warranty cards for the furnace. Each one has a date that matters, and no family has a system for tracking all of them.

The Search Tax

There's a less obvious cost: the time spent looking for things. You need your T4 slip from two years ago for a mortgage application. You know you have it. You remember scanning it. But was it in the "Tax" folder? The "2024" folder? Did you email it to yourself?

A 2023 study from the Paperless Project found that the average person spends 18 minutes per document search. For a family dealing with tax season, a medical claim, or a home purchase, that adds up to hours of searching per month. Not because the documents are lost -- because they're filed in ways that made sense at the time but don't match how you actually need to find them later.

This is the gap between storage and retrieval. Google Drive, Dropbox, iCloud -- they all solve storage. None of them solve retrieval when you can't remember the file name.

What We Decided to Build

When we started Archevi, the insight wasn't "families need another place to put files." It was that families need to ask questions and get answers from the documents they already have.

Upload your auto insurance policy as a PDF. Later, ask: "When does my car insurance expire?" Archevi reads the document, finds the relevant clause, and tells you -- with the source highlighted so you can verify it yourself.

The difference is semantic search. Instead of matching file names or keywords, Archevi understands what your documents are about. A search for "property tax deadline" will find your municipal tax assessment even if the file is named "2025_city_notice.pdf" and the word "deadline" never appears in the document.

We use Cohere's embedding model to convert documents into vector representations, stored in PostgreSQL with the pgvector extension. When you ask a question, your query is embedded the same way, and the system finds the most semantically similar document sections. It's the same technology that powers enterprise knowledge management -- but configured for the scale and privacy requirements of a family.

The Privacy Tradeoff We Made

Family documents are sensitive. SIN numbers, medical records, financial statements -- this isn't data that should travel through third-party services without protection. We run a privacy layer that anonymises personally identifiable information before it reaches any cloud AI service. Your name, your SIN, your address -- these are replaced with surrogates during processing and restored in the final answer.

Is this approach perfect? No. Anonymisation adds latency, and the privacy layer occasionally over-redacts common words. We decided the tradeoff was worth it: slower responses are acceptable; data exposure is not.

All document storage runs on Canadian infrastructure -- a DigitalOcean server in Toronto. Not because Canadian hosting is inherently more secure, but because it keeps your data under Canadian privacy law (PIPEDA) rather than the US CLOUD Act, which allows American authorities to compel access to data stored by US companies regardless of where the server sits.

What This Looks Like in Practice

A family of four uploads their insurance policies, tax returns, medical records, and household warranties. That's roughly 40-80 documents in the first month.

From that point forward:

• "How much did we pay for home insurance last year?" -- answered in seconds, sourced from the actual policy document.
• "What's the warranty status on the dishwasher?" -- found by content, not by remembering you named the file "KitchenAid_W10632081A.pdf."
• "What medical expenses can we claim this year?" -- cross-referenced across multiple receipts and the CRA's eligible expense list you uploaded.

The annual value isn't dramatic -- we estimate $300-700 in avoided costs (missed renewals, unclaimed deductions, duplicate services) plus 15-20 hours of search time saved. At $72/year for the Starter plan, the math works.

The Honest Limitation

Archevi doesn't replace a financial advisor, an insurance broker, or an accountant. It's a retrieval system, not an advisory one. When it tells you your insurance expires on March 15th, it's reading your policy -- not recommending whether to renew.

The value is in the retrieval: getting the right information from your own documents at the moment you need it, without spending 20 minutes searching first. For most families, that's the part that was broken.