Case study · Built by our founder
College Cost Secrets
A web app that connects families with admissions and financial professionals. Families onboard, get their federal Student Aid Index calculated, then search, favorite, and compare colleges side by side, while agents, admission professionals, and admins each work from their own dashboard.
The story
College Cost Secrets predates Inseed. Our founder built it as a frontend engineer on Rehan's LaunchX engagement, years before Inseed existed. It is founder-portfolio work, not an Inseed client project, and it is the story the team leans on for complex, multi-role web apps with real domain logic behind them.
The problem it tackles is a genuinely hard one to make simple. Families need to understand what college will actually cost them, what aid they might qualify for, and which schools give their student the most free money. The professionals advising them, agents and admission professionals, need their own tools to onboard families, run the numbers, and manage a caseload. Four different kinds of user, all looking at the same underlying financial picture from a different seat.
On top of that sits real math. The federal Student Aid Index (previously the Expected Family Contribution) is not a marketing estimate, it is a defined federal calculation driven by household income, family size, and tax situation. The product had to compute it correctly and then turn it into a plan a family could act on.
What we built
The build is a role-based web application with distinct flows for families, agents, admission professionals, and admins, all reading from the same core data. A family signs up, walks a multi-step onboarding wizard (household, parents, household finances, students), and comes out the other side with their federal Student Aid Index and a financial plan, rather than a data-entry form that leads nowhere.
From there the family works the college-search side: filtering schools by admission fit, location, cost, and major, favoriting them, saving searches, and opening a per-college report with a full academic and financial breakdown and side-by-side comparison. Admission professionals and agents invite and manage families from their own portals, and an admin console sits over the top managing agents, professionals, and families.
Our founder owned the frontend. The backend was a separate service the founder's frontend consumed, so the case study is honest about that split: this is a proof point for building a large, multi-role, data-heavy web frontend with real financial logic surfaced through it, not a claim of owning the backend.
Architecture
The web app is a Next.js and React build in TypeScript. Client state runs through Redux Toolkit, server state through React Query, and both talk to an external REST API over axios. Sessions use Firebase behind a next-auth layer, Tailwind handles styling, and Formik with Yup drives the heavy multi-step onboarding wizard.
Around that core sit the pieces that make the product feel researched rather than estimated: Algolia powers the college search, Nivo charts and react-simple-maps render the dashboards and per-college reports, and react-joyride guides first-time families through the flow. Sentry and Google Analytics watch production, and the web layer ships on Vercel. The backend it consumes is a separate REST service, so the app was built against that boundary rather than coupled across it.
Selected screens
Stack
One decision worth telling
Most "college savings" tools stop at a vague number. This one had to compute the federal Student Aid Index for real, from household income, family size, and tax situation, and then split it into the institutional and federal figures that actually drive aid eligibility.
That put a defined, changeable federal formula at the center of the product instead of at the edge. The onboarding wizard exists to gather exactly the inputs that calculation needs, and the college-search and per-college reports downstream all read off the resulting figure. Getting that calculation right, and presenting it in plain language, was the load-bearing piece the rest of the experience hangs on.
Outcome
What we would do differently
Most agencies skip this section. We include it because honest hindsight is the kind of credibility you cannot buy.
The product carried two identities across the app, an earlier working name alongside the College Cost Secrets branding, which shows up as inconsistency between screens. Locking the name and a shared component library up front would have kept every portal visually on the same page and cut the rework of restyling later.
The Student Aid Index is federal policy, and it changes (the EFC-to-SAI transition is the obvious example). Keeping that math in one versioned, well-tested module, rather than threaded through the UI, would make each annual formula change a contained update instead of a hunt across screens.
Family, agent, admission professional, and admin each grew their own flows and dashboards. Treating roles and permissions as a first-class part of the data model early, instead of extending per portal, would keep the four experiences from drifting apart as features land.
Related work
Built by Inseed
Zero to App Store in under 4 months. Past 3,000 users in the next 3.
Read the case studyBuilt by Inseed
Vibe search shipped in 3 weeks. AI searches up ~200%.
Read the case studyRole-based dashboards, real domain math, search and comparison, and the data viz to make it legible. Tell us what you are building and we will tell you what the first four weeks can de-risk.
Book your audit