Project 04 · Group Seat Finder · demo

How do you seat four friends who all want different things?

Booking West End tickets for a group is a negotiation. I turned it into a maths problem, then into a tool.

Focus: The maths

Amol

Cheap seats

Ashley

Good mix

Ram

Up close

Rohnit

Good mix

Prototype data, modelled on a West End theatre

Step 1 of 5

Four friends want to see a show.

Amol wants it cheap, Ram wants to see the actors up close, and Ashley just wants to sit with Rohnit.

Step 2 of 5

First, the price of every seat.

Scraped from the real booking funnel. Orange is expensive. Grey is already gone.

Step 3 of 5

Then, how good each seat actually is.

Not from the seller. Estimated from geometry: distance to the stage and how central you are.

Step 4 of 5

Everyone gets a slider.

Each person moves a slider for what matters more to them: saving money or a better view. The tool then scores every seat for each of them.

Step 5 of 5

One row, four seats, everyone happy.

It searches every contiguous block and picks the one with the highest total happiness. Optimal, not greedy.

Role: Solo build
Algorithm: Hungarian matcher
Data: Headless scraper
Status: Prototype

Built with TypeScript · React · Hungarian algorithm · Headless browser

Next project

Portfolio Dashboard →

Pulls statements from my inbox, reads the PDFs, merges broker holdings and charts it all. Personal financial data never leaves my laptop.

Setting the type…

The Yash Hour

Project 04 · Group Seat Finder · demo

How do you seat four friends who all want different things?

Booking West End tickets for a group is a negotiation. I turned it into a maths problem, then into a tool.

Focus: The maths

Amol

Cheap seats

Ashley

Good mix

Ram

Up close

Rohnit

Good mix

Prototype data, modelled on a West End theatre

Step 1 of 5

Four friends want to see a show.

Amol wants it cheap, Ram wants to see the actors up close, and Ashley just wants to sit with Rohnit.

Step 2 of 5

First, the price of every seat.

Scraped from the real booking funnel. Orange is expensive. Grey is already gone.

Step 3 of 5

Then, how good each seat actually is.

Not from the seller. Estimated from geometry: distance to the stage and how central you are.

Step 4 of 5

Everyone gets a slider.

Each person moves a slider for what matters more to them: saving money or a better view. The tool then scores every seat for each of them.

Step 5 of 5

One row, four seats, everyone happy.

It searches every contiguous block and picks the one with the highest total happiness. Optimal, not greedy.

Role: Solo build
Algorithm: Hungarian matcher
Data: Headless scraper
Status: Prototype

Built with TypeScript · React · Hungarian algorithm · Headless browser

Next project

Portfolio Dashboard →

Pulls statements from my inbox, reads the PDFs, merges broker holdings and charts it all. Personal financial data never leaves my laptop.