Shoham Fellner

Case Study

OctSeven

October 7th Memorial Platform · Volunteer · 2023-2026

A public platform where families build memorial pages for the people they lost. Development started two weeks after October 7th.

RoleCore Frontend Developer

Stack

[Next.js] [Redux] [MUI]

Visit Site

Home page - memorial grid

Live Site

octseven.com is live and public. Shown here are the grid page and a blank creation form - deliberately not an individual memorial page, out of respect for the families whose stories they hold.

The Stakes

Built for the families and friends of the people lost on October 7th, so they could create and share memorial pages for the loved ones they miss - a volunteer initiative, not a company product. Every profile on it represents a family who trusted us with someone they lost.

Being volunteer-built doesn't lower the stakes here - the subject leaves no room for the usual margin of error. A bug that mixed up two people's data, or lost a family's story, wouldn't just be a defect. It would hurt people who chose to hand us these memories directly.

My Role

Part of a fully volunteer team from the project's first day to its last, alongside a rotating cast of other developers, plus product, UX/UI, UX writing, QA and DevOps - some there for the duration, some for a stretch. Nobody on the team had worked together before, and we were spread across time zones and technical backgrounds.

I built most of what a visitor actually touches: the home page's profile grid and filtering, the create-profile form, the add-story form, and the profile page itself. Next.js was already chosen before I joined, for its SEO benefit on a public memorial site.

The Problem

The team formed to get a platform live as close to October 7th as possible - strangers, remote, from different technical backgrounds, building toward one deadline with no slack in it.

Speed was the actual constraint, and it didn't let up: every technical choice had to hold up under a team that kept changing shape while the work was still in motion.

Decisions

A person checks every profile before it goes live

Constraint: A mix-up on a memorial profile isn't a minor bug - it's someone's trust in what we built with their loved one's memory.
Decision: A manager reviews every new profile for duplicates and accuracy before it publishes, and a profile's own author has to approve any story someone else adds to it.
Trade-off: Nothing goes live instantly - every submission waits on a person, which is slower than an open pipeline, and the right trade on a project where a wrong page is worse than a slow one.

Forms built from a schema, not hand-coded per case

Constraint: The create-profile and add-story forms needed to hold up while the team itself kept changing, without every new contributor relearning bespoke form code.
Decision: Schema-driven forms - the same technique later projects would also lean on.
Trade-off: Some indirection between what a form shows and where that's defined - a fair trade against a rotating team that couldn't afford to onboard onto one-off form logic each time someone joined.

Reaching for the tools everyone already knew

Constraint: A team of strangers, on a deadline, with people rotating in and out mid-project.
Decision: Redux and MUI over less familiar alternatives, chosen because they're the industry-standard tools most incoming contributors would already recognize.
Trade-off: Neither was necessarily the best technical fit on its own merits - the fit that mattered here was a shared team already knowing them.

Evidence

octseven.com is live - what follows is the grid and a blank form instead of a real memorial page, to protect the families and keep their loved ones' memory in their hands, not a case study.

The full memorial grid, with filtering.

The schema-driven form used to create a profile.

Outcome

Shipped and live at octseven.com, built and launched by a volunteer team that had never worked together before, on a timeline measured in weeks.

The platform is feature-frozen now, and will be retired soon - its data merged into another existing platform, a maintenance-cost decision rather than a product one.

What I'd Change

The code carries real duplication across the create-profile and add-story forms - a direct cost of a volunteer team that kept changing shape, where consistency across contributors was harder to hold onto than the features themselves. I'd go back and refactor it properly.