Rohan Jasani

I find where a team’s week is going, then build the AI that gives it back.

I’m Rohan, a solution designer who builds. I sit with the people doing the work, find what is quietly eating their time, then design the fix and build it myself. Most recently that was the AI operations platform behind a mortgage brokerage in Vancouver.

Based in Vancouver. Open to applied AI roles across Canada.

A three-minute tour of the platform I built at Level Up. Everything on screen is sample data.I made the tour with Margin, a tool I built for turning a script into a product walkthrough.

It started with one question

Level Up is a small mortgage brokerage. When I arrived for a summer internship from UBC, nobody handed me a spec. So I asked the owner and the operations lead one question: where is the week actually going?

Then I followed a deal through the office to find out. A lot of skilled time was going into chasing rates, sorting paperwork, and working out which documents were still missing. None of that was the job. All of it still had to be right.

What I built in response is Level Up’s internal AI operations platform. The team calls it Lum Brain. It connects the tools the brokerage already relied on, including Gmail, Zoho CRM, Finmo, Google Drive and Slack, and takes over the connective work between them.

Level Up’s public page for the platform. A sample lender rate sheet sits beside the answer the assistant drew from it.
I also designed and built the public page that explains the platform to Level Up’s clients. It runs on sample data. Visit levelupmortgages.ca.

Three things it took off the team’s desk

Rates
Lenders email their rate sheets, and somebody used to retype them. Now each lender’s rates, promotions and policies keep themselves current: 66 lenders, more than 1,200 updates.Counted from the production database on October 5, 2026: 66 lender rate cards and 1,267 revisions. I ran totals only. No client details were read.
Paperwork
More than 3,900 client documents read, named and filed to the right folder in three months.Counted on October 5, 2026: 3,925 documents since July 10. On filing alone, that gave back an estimated 8 to 12 staff hours a week.This one is an estimate, and it covers filing only. It assumes 1.5 to 2.5 minutes to download, rename and upload each document. The total across the platform is higher.
Chasing conditions
A deal planner that looks like the spreadsheet the team already trusted, and knows which of a lender’s conditions have their documents. The first version looked nothing like a spreadsheet, and nobody used it.
The full story, a 7-minute read Rates, paperwork and a spreadsheet nobody wanted to give up How each piece works, what I got wrong first, and where a person still makes the call. Read the story

The platform is still running: 99 scheduled automations, more than 145,000 runs, a 99.1% success rate.Also counted on October 5, 2026: 145,394 runs, 1,300 failures. I still support the team part-time as new needs come up.

How I work

The technology mattered less than the loop behind it. It has four steps, and I keep going around until people stop noticing the tool.

One lap at Level Up: the deal planner. The first version was cards, and nobody used it. The one the team kept looks like their spreadsheet.
  1. I learn a business by sitting with the people who run it. The problems worth solving rarely make it into a requirements document. They turn up when someone says, “and then I have to…”

  2. A working thing within days tells you more than a plan does. Small also means that when it is wrong, it is obvious and cheap.

  3. Usage is the honest feedback. If the team quietly works around a feature, the feature is the problem.

  4. Most of the value shows up after launch, in the unglamorous fixes that turn a tool into something people rely on without thinking.

I do the design and the building myself, working alongside AI coding tools like Claude Code and Codex. That means the person who understood the problem is also the person who ships the fix, and nothing gets lost in translation between the two.

At Level Up that made me the analyst, the designer, the engineer, the help desk and the technical writer, often in the same week. I liked it that way.

Before this

  1. 2016 to 2021

    University of Utah

    Bachelor of Science in Business Management, with a minor in Entrepreneurship. Innovation Scholar.

  2. 2022

    Megaplex Theatres, Salt Lake City

    Data Analytics Intern. I built a Python pipeline to test whether YouTube trailer views predict ticket sales. They barely do, which was the useful answer.

  3. 2022 to 2023

    Goldman Sachs, Salt Lake City

    Consumer Wealth Management Analyst. I reviewed more than 200 onboarding packages a week for high-net-worth clients, flagged the ones that didn’t add up, and worked with sales and operations to close 95% of documentation gaps within 48 hours. It is where I learned that trust is mostly details.

  4. 2025 to 2026

    UBC Sauder School of Business

    Master of Business Analytics. Machine learning, optimization and decision-making under uncertainty, all pointed at business problems.

  5. 2026

    Level Up Mortgages, Vancouver

    Applied AI Solutions Engineer (Intern), May to September 2026. The work at the top of this page.

Work Study In order, not to scale.

Things I built because I wanted them to exist

Margin
A scene from a walkthrough made with Margin, with a caption along the bottom. Turns a written script into an animated, captioned product walkthrough. An AI agent drafts each scene and times the captions. I built it to explain the Level Up platform to people who were never going to read the documentation. Watch a tour made with it.
Breath Box
Breath Box: a soft orange orb inside a ring of beads that light up one by one. A box-breathing wallpaper for the desktop, the phone and the new-tab page. An orb, a ring of beads, and no words. I built it for my own screen, and it is open for you to use: what you write there stays in your browser. Breathe with it.
Career Constellation
Career Constellation: a map of 622 job descriptions as coloured dots grouped into job families, beside a list of findings. A hackathon build on Methanex job data. It takes messy, inconsistent job titles and turns them into a map of roles and the paths between them. Explore the map.
ALGS Dashboard
The ALGS Dashboard: one bar for each team, built from a block for each game in the colour of its map, beside a short account of the last game. Results from the Apex Legends Global Series, made explorable. The standings replay game by game, so you can watch a tournament turn. It found an audience on Reddit. Open the dashboard.
The Kitchen
My setup for working with AI coding agents, modelled on a restaurant kitchen: one plans, others cook, a fresh one tastes. Its main rule is that every correction gets fixed at the strongest layer available, so the same mistake can’t happen twice.

About me

I’ve always worked in the gap between two kinds of people: the ones who know the problem and the ones who know the technology. Most teams have that split. I like standing in it and closing it.

I came to the US to study video game development and ended up with two business degrees. Games taught me how systems behave. Business taught me which problems deserve a system at all.

People ask whether I’m a designer, an engineer or an analyst. Mostly I’m the person who asks the first question. Where is the week going? Who touches this twice? What happens if we just stop doing it?

The answers usually point at something I can build.

This is Marvin. He automates nothing and saves nobody any time. I built him because it was fun. Poke him a few times, or pick him up. Tap him a few times.

A few things I believe

  • Useful beats impressive.
  • AI should suggest. People should decide the things that matter.
  • If a team has to be talked into using a tool, the tool isn’t finished.
  • When a number is an estimate, say so.

If your team’s week is disappearing somewhere, I’d like to hear about it.

Email is the fastest way to reach me. I’m in Vancouver, authorized to work for any employer in Canada, and open to relocating within it. I speak English, Hindi and Gujarati.

My GitHub looks quiet this year because the platform lives in Level Up’s private repositories. I’m happy to walk through the code on a call.