DEV Community: Rozeena

Graph Hacks build log: I gave my dealership AI a memory for its own alerts

Rozeena — Thu, 15 Oct 2026 08:54:35 +0000

I keep building Dealership Brain and finding the same embarrassing failure mode: the app nags dealerships about missed callbacks, but had no idea whether its own alerts ever actually went out. Alerts fired into the void. If something failed, nobody knew — including me.

So today, during the Graph Hacks window, I gave it a memory.

What's new

Persistent Alert Log — every n8n alert attempt is now recorded with a status: sent, preview, or failed. Each entry stores what triggered it (usually an overdue callback), when it fired, and the outcome. No more mystery about whether the workflow actually ran.

Manager Alerts tab — managers get a dedicated tab in the UI showing the alert history. It's the first thing I check now after a busy morning: did the system actually flag everything it should have?

n8n alert button for overdue callbacks — there's now an explicit trigger that fires the n8n workflow when a callback crosses its SLA deadline. Before this, the overdue check and the alert lived in different places and occasionally disagreed.

Why this matters

An AI that reminds humans to follow through on promises needs to be held to the same standard. If the app forgets to send an alert, that's the exact failure it's supposed to prevent. Eating my own dog food, basically.

The honest state of it: the log is simple, the tab is unstyled, and the failure statuses need better retry handling. But the core loop works — promise missed, alert fired, attempt recorded.

Try it: https://dealership-brain.vercel.app
Code: https://github.com/rossycoder/dealership-brain

What Will Your City Feel Like in 2050? I Built a Climate Detective With 40 Years of NASA Data

Rozeena — Wed, 14 Oct 2026 07:45:35 +0000

Ask a weather app about climate change and it shows you a chart. Charts don't scare anyone.

So TerraDetective AI answers differently: your city in 2050 will feel like a city that exists today. Karachi's 2050 climate twin? Riyadh, now. That lands differently — because you've seen Riyadh's heat. You can picture living in it.

I built TerraDetective for the NASA Space Apps Challenge on 40 years of real NASA POWER data. No projections invented by a language model — measured data, analyzed live.

The feature to try first: Future Twin City. Pick your city; the tool analyzes its temperature trajectory and finds its 2050 climate twin among real cities today.

Also inside:

Was This Normal? — pick any month in history, get a z-score verdict: normal, unusual, or off the charts
Danger Days — count of days per year above heat thresholds, trended over decades
Real statistics — every trend ships with a 95% confidence interval and a p-value. If the trend isn't significant, the tool says so.

That last point is the whole philosophy: a climate tool that overstates its case is just propaganda with better charts. Every number carries its uncertainty.

Run your city through it:
👉 https://terradetective-ai.vercel.app
Code: https://github.com/rossycoder/terradetective-ai

Built for NASA Space Apps. Tell me your city's 2050 twin in the comments — the surprising ones are the best argument for building tools like this.

Smog Season Is Coming — I Built a Free Tool That Tells Parents If Kids Should Go to School

Rozeena — Wed, 14 Oct 2026 05:57:20 +0000

In Lahore and Delhi, winter smog season means the same morning question in millions of homes: is it safe to send the kids to school today?

AQI apps answer with a number. Numbers don't make decisions — parents do, at 7am, half-awake. I built HawaGuard for the Code for a Billion hackathon to turn air-quality data into decisions.

🛡️ Start with Family Shield: enter your child's age and any respiratory sensitivity, get a plain school-day verdict — go, go with a mask, or keep home if you can — with the reason stated, not just a color.

Also built in:

Outdoor Planner — for walkers and the elderly: not just "AQI now," but the best hour today to go out
Symptom Checker — is this cough pollution irritation or an infection worth a clinic visit? (Guidance, not diagnosis — it says so in the app.)
Roman Urdu mode — full guidance in the language people actually speak at home

Decisions need local data homeowners trust, so HawaGuard uses labeled monitoring stations. A note on honesty: one station series in the demo is Beijing historical data, clearly labeled — because PM2.5 health physics is the same everywhere, and pretending it's local when it isn't would break the whole point of the app.

Check today's verdict for your city:
👉 https://hawaguard.vercel.app
Code: https://github.com/rossycoder/hawaguard

Parents — what do you actually check before school on a smog day? Building the next version from your answers.

CodeForABillion #PublicHealth #AirQuality #Smog #Hackathon #buildinpublic

I Built a Free Tool That Shows Small Shops Exactly How Much Cash Is Trapped in Dead Stock

Rozeena — Wed, 14 Oct 2026 05:56:54 +0000

A small shop owner I know had thousands of dollars of stock sitting on shelves that hadn't moved in months. Still paying rent on that shelf space. Still reordering the things that DID sell — by gut feel.

Retail doesn't have a data problem. It has a trapped-cash problem. I built StockSentry for the InfiniSynapse Vibe Coding Hackathon to put a number on it — free, no signup, works from a CSV export your POS already makes.

Start here: the Dead Stock Detector. Upload your inventory and it tells you exactly how much cash is locked in SKUs that haven't sold in 30/60/90 days. Not a chart. A number, in dollars, per item.

Then:

Discount Ladder Simulator — if you discount 10/20/30%, how long until it clears? What do you recover vs. write off?
Viral Spike Simulator — a product goes viral overnight. What breaks first? What do you lose by stocking out on day two?
True Restock Calculator — reorder points from actual velocity, not habit

One promise on the tin: StockSentry labels every number honestly. If a figure comes from a heuristic, it says so. If it's AI-assisted analysis, it's badged as such. Black-box "AI says so" is how tools lose trust with real shop owners.

Try it with the built-in demo CSV:
👉 https://stocksentry-bice.vercel.app
Code: https://github.com/rossycoder/stocksentry

Shop owners and retail folks: what's your dead-stock number? I genuinely want to know if the detector matches what you see on your shelves.

smallbusiness #retail #inventory #ai #hackathon #buildinpublic

I Built an AI "Brain" for Car Dealerships That Never Forgets a Customer

Rozeena — Tue, 13 Oct 2026 08:20:17 +0000

A customer calls a dealership and is promised a callback in 15 minutes. Nobody calls back. They buy from a competitor.

This isn't a sales problem — it's a memory problem. Salespeople forget. CRMs don't remind. So I built Dealership Brain for the FalkorDB Graph Hacks hackathon: an AI operations brain that stores every customer, every promise, and every conversation in a knowledge graph — and escalates before a broken promise costs a sale.

🔴 The feature that matters most: Callback SLA Escalation. Every callback promise gets a timer. Miss it, and the case automatically escalates to the manager. No customer silently falls through the cracks again.

Also inside:

Promise Tracker — every commitment made to a customer, tracked to done
Smart Match — a new car arrives, the brain finds buyers who wanted exactly that, and drafts the WhatsApp message
Follow-Up Agent — surfaces who to contact today and why
Trade-In Valuation — trade-in value + remaining loan → net equity
Test Drive Booking — double-booking is rejected outright
Neighborhood Graph — customers visualized as a connected network
Roman Urdu chat — because that's how many customers actually write
Staff Leaderboard + service reminders

How it works: a FastAPI backend over a graph store. Every interaction becomes nodes and edges, so the system reasons about relationships, not rows. Ask "which customers were promised a callback this week and never got one?" — a graph answers that natively. A traditional CRM makes you dig.

Try it — simulate a missed callback and watch the escalation fire:
👉 https://dealership-brain.vercel.app
Code: https://github.com/rossycoder/dealership-brain

Built for FalkorDB Graph Hacks. Feedback welcome — especially from anyone who's worked a dealership floor: what's the promise your team breaks most?

buildinpublic #ai #hackathon #python #webdev