Case Study · AI Product

Smart Home Digest

AI · Scoring · Telegram

Homebuyers get 200+ listing alerts a day. Most are noise. I built an AI that reads a buyer's real priorities and delivers the 3–5 homes that actually matter — with the reasoning visible.

← All projects

The Problem

Zillow and Redfin are built to surface everything. That's great for the platforms — more alerts means more engagement — but it's the wrong product for a buyer who's already overwhelmed. The signal is buried in the noise.

Real estate agents solve this with judgment. But agent judgment isn't available at 7am, isn't personalised to your exact criteria, and isn't scalable. I wanted to see if AI could replicate that filtering instinct — and be transparent enough that buyers would actually trust it.

Users & What I Learned

Primary User
The homebuyer

Wants to open one message in the morning and know exactly which 3–5 homes to look at. Not a ranked list of 50. Not a spreadsheet. A confident recommendation they can act on.

Critical Insight
Trust is a product feature

Test buyers override a black-box ranking. They need to see why a home scored 85 vs 45 — which factors drove it, which deal-breaker penalised it. Without transparency, the system gets ignored.

This insight shaped the entire algorithm design. Transparent scoring with visible factor weights wasn't a nice-to-have — it was the thing that made the product usable. A buyer who understands the score acts on it. A buyer who doesn't, doesn't.

Decisions I Made and Why

Must-haves at 25%
Non-negotiables get the highest single weight.

User research showed buyers stretch on budget before they compromise on a must-have. Weighting must-haves above price reflects how buyers actually make decisions — not how platforms model preferences.

Delivery Channel
Telegram, not email.

Email digests get archived unread. Telegram gets opened. Channel choice is a product decision — the best ranking algorithm fails if nobody sees the output. I tested both and measured open rates.

Deal-breaker Penalties
Negative scoring, not exclusion filters.

Hard exclusions remove homes buyers might want to see with caveats. A −5% penalty per deal-breaker keeps them visible while clearly signalling the tradeoff — and buyers can make the final call.

Outcomes

2
real buyers tested
with live profiles
7
weighted
scoring criteria
Daily
automated delivery,
zero manual steps
0–100
transparent score
with visible reasoning

The Scoring Algorithm

Must-haves
25%
Budget fit
15%
Location
15%
Basics (bed/bath/sqft)
15%
Nice-to-haves
15%
Schools
10%
Deal-breakers
−5% ea

5-Agent Processing Pipeline

Behind every ranking is an AI pipeline that fetches listings, normalises data, scores homes intelligently, and formats the digest for delivery.

📋
Buyer Profile Form input
0
Fetch Listings Redfin scraper
1
Normalize Standardise JSON
2
Score & Rank 0–100 w/ reasoning
3
Format Digest Telegram delivery
Pipeline: User submits buyer profile via form → Agent 0 scrapes Redfin with filters → Agent 1 normalises raw listing data to standardised JSON → Agent 2 scores each home 0–100 with transparent reasoning → Agent 3 formats as warm Telegram digest → Delivered daily with zero manual steps.

Try the Demo

Fill in your profile to see how the algorithm ranks Bay Area homes. All data is dummy data for demonstration.

Budget Range

Preferred Locations

Must-Have Features

Nice-to-Have Features

Tell us about other features that matter to you

What This Demonstrates

🎯 Algorithm Design

Not guessing. Transparent weighting: budget 15%, location 15%, must-haves 25%. Every factor has a reason tied to user needs.

👥 User Research & Validation

Tested with real buyers. 2 test profiles with different budgets and priorities show the system works across personas.

🔍 Transparent Logic

Users see exactly why a home scored 85 vs 45. Builds trust. Shows you think about user confidence, not just accuracy.