Why I built Bridge

I ran a bike shop on Lightspeed Retail for fifteen years. Long enough to know exactly where every answer I needed was hiding — and exactly how long it took to dig it out.

That gap is the reason Bridge exists.

The Saturday-night question

It's the end of a good Saturday. You want one thing before you lock up: did we actually beat last year, or did it just feel busy?

That's the first question. It's never the only one. The second you have that number you want the next — was last week's weather better or worse than the same week a year ago, and is that what moved it? How many work orders did we take in this week versus the same stretch last year, is service keeping pace with the floor? And the bikes we did take in — how fast are we turning them around, from the minute they hit the rack to the minute the customer rings out?

Every one of those is answerable. Most of it is sitting in Lightspeed; the weather is a thirty-second search away. But getting them lined up means opening reports, setting a date range, setting the same range a year back, squinting, and switching tabs. Five minutes if you do it often. Longer if you don't, because you half-remember which report has the comparison you want.

So most nights you don't bother. You go home with a feeling instead of a number. Multiply that by a few hundred decisions a year — what to reorder, who to call back, what's quietly dying on the shelf — and you start running the shop on instinct when the data was right there the whole time.

Fifteen years of pulling the same reports

That was the job. Not the fun part of retail — the part underneath it.

Which items haven't sold in ninety days. Who my top customers were this quarter and what they actually bought. Which service customers were overdue to come back in. Every one of those is a question with an answer sitting in the database. And every one of them was a report to find, run, export, and read.

The information was never the problem. The friction of getting it out was.

The plug-in graveyard

I tried to fix this the way everyone does. I bought things.

Over the years I paid for dozens of plug-ins and add-ons that each promised to solve one piece of it. Most did, sort of. And every one had a hard limit — on the reporting, on what you could actually ask, on what shape your business had to fit into to use it. You'd get the one report it was built for, and the moment you wanted to ask the next, slightly different question, you were back to exporting to a spreadsheet.

None of them ever gave me the thing I actually wanted, which was simple: to ask a normal question in plain English and get a straight answer from my own numbers.

What changed

Two things, close together.

AI assistants got genuinely good at answering questions in plain language. And a standard called the Model Context Protocol — MCP — became the common way to let those assistants talk to other software. Anthropic introduced it in late 2024; by the end of 2025, every serious assistant supported it.

That combination is the whole opportunity. Your assistant is already good at the conversation. MCP is the part that lets it reach your data. The missing piece was a connector built for your point-of-sale, by someone who knew the workflow.

So instead of building another dashboard for you to read, I built the layer that lets the assistant read the store for you.

What Bridge is — and what it isn't

Bridge, by Rizolvr, is a hosted MCP server for Lightspeed Retail. You connect it to your Lightspeed account once. From then on, the AI assistant you already use — Claude, or anything that speaks MCP — can answer questions about your store from live data. Sales, inventory, customers, work orders, purchase orders.

Now that Saturday-night question is a sentence. So is the next one, and the one after that:

• "How did we do this Saturday versus the same Saturday last year?"
• "How many work orders did we take in this week versus the same week last year?"
• "What's our average turn time right now, from check-in to ring-out?"

And because it's your assistant — the same one you use for everything else — you can hand it the thing Lightspeed doesn't even know: "Now factor in how last week's weather compared to the same week a year ago." It pulls the weather itself, lays it next to your numbers, and tells you whether the rain explains the dip. No report to find. No export. No second tab.

I'll be straight about the boundary, because overselling this would be the fastest way to lose you. Today Bridge reads your Lightspeed data — it doesn't change it. It answers from your most recent sync, not live to the second. And it connects Lightspeed Retail, not every POS under the sun. That's the honest shape of it today.

But I'll tell you where this is headed, flagged as exactly what it is — not shipped yet. The next version of Bridge doesn't just read your store. It writes to it. Photograph a supplier's packing slip and tell your assistant to build the purchase order — it builds it. Hand it the invoice when the costs come in different and tell it to update your pricing — it updates it. The back-office shuffle turns into a conversation. That version is in final testing right now — and that's the part where Bridge stops being a faster way to look things up and starts being a different way to run the shop.

That experience — fifteen years of knowing the answer was in there and having to go dig it out — is the reason Bridge looks the way it does. It's not trying to be a platform. It's trying to get out of your way and let you ask.

If you run a shop on Lightspeed, see what it costs.