SLICs
Instant destination lookups for UPS Feeder drivers — no more digging through a binder.
[Live demo link coming soon] · [https://github.com/WebDevilishSmile]
The Problem
Before a Feeder driver leaves the building, dispatch hands them a SLIC — a short code identifying their destination (e.g. 1809). To actually get there, the driver then had to either flip through a printed binder of addresses or ask dispatch to look it up and print a slip with the address and phone number. It worked, but it was slow, paper-dependent, and repeated the same manual lookup hundreds of times a week across the building.
What It Does
SLICs replaces the binder with a single search bar. A driver types (or scans) a SLIC and instantly gets:
- The destination's address and phone number
- A one-tap link that opens the address directly in Google Maps
- A tappable phone number that calls the destination's dispatch right from the app
- Driver-added comments on each SLIC, so drivers can share firsthand knowledge that never makes it into an official address — where the actual entrance is, which roads to avoid (height/weight restrictions), who to contact on-site if there's an issue, and other details you only learn by having been there
- An admin panel (built for myself, as the maintainer) to add, edit, or remove SLICs as locations and details change — no redeploy needed
Real-World Impact
This isn't a demo project — it's used daily by drivers in a building of roughly 270. A few numbers from production analytics:
- 200 of 269 drivers in the building have used the app at least once since it launched in early 2025
- Averaging roughly 130 combined lookups/day across both live deployments as of September 2026
- Individual destination pages see far more repeat visits than unique visitors over the same period — drivers are coming back to the same lookups again and again as part of their regular routes, not just trying it once
(Numbers pulled from Vercel Analytics + app database — update periodically rather than treating as fixed.)
Tech Stack
- Framework: Next.js / React / TypeScript
- Database: MongoDB
- File storage: Vercel Blob (per-destination directions PDFs)
- Auth: Google sign-in / Email/password
- Access: open sign-up, decided on purpose (2026-10-28): anyone can make an account and use the SLIC lookup, tips and places, with no approval step. Email/password accounts get a link to confirm their address, which is what lets the app trust it (a Buy Me a Coffee membership, linking a later Google sign-in). The weekly cover bid (Covers) has its own gate: an admin links the account to the driver's name on the on-call sheet.
- Hosting/Analytics: Vercel
Notable Engineering Details
- Admin CRUD without redeploys — location data changes constantly (new stops, updated phone numbers), so the admin panel lets data stay current without touching code.
- Fast search over a growing dataset — built to return a match instantly even as the number of stored SLICs has grown.
- Crowdsourced local knowledge — official address data can't capture a hidden entrance, a bridge with a low clearance, or who to actually call when a dock is closed. Letting drivers comment directly on a SLIC turns the app from a static directory into a shared knowledge base built from what drivers have actually experienced on that route.
Screenshots
[Coming soon — screenshots of the search bar, a destination page, and the admin panel.]
Why I Built This
I'm a Class A Feeder driver — this wasn't a coding exercise, it was a problem I ran into on the job every day. I built it, shipped it, and kept improving it based on what other drivers in my building actually needed.
