How Helping My Son Find an Apartment Sparked a Startup
A founder’s story about the problem hiding in plain sight - and why AI finally makes it solvable.
Last summer, my son graduated from the University of Pittsburgh and landed a finance rotation program at a major bank. We were thrilled. What we weren’t prepared for was the apartment hunt.
For three weeks, my kitchen table became mission control for the most disorganized operation of my career - and I’ve been a software engineer for over two decades.
We had 40 browser tabs open across Apartments.com and Zillow. A shared Google Sheet that was always out of date. Phone photos we couldn’t remember which apartment they belonged to. Email threads with landlords that disappeared into our inboxes. And a running text chain that consisted almost entirely of “wait, which one had the parking fee?” and “did you hear back from the one on Market Street?”
We toured six apartments in three days. By the end, every apartment had merged into a single beige blur. We made a decision based mostly on exhaustion. It worked out fine, he got a decent place, but the process was inexcusable.
Here’s the thing that nagged me: I build software for a living. I’ve shipped products used by millions of people. And I just spent three weeks doing a task that should have taken three days, using tools that haven’t meaningfully improved since Craigslist.
The Problem Hiding in Plain Sight
Apartment hunting isn’t a search problem. Zillow and Apartments.com solved search years ago. The real problem is everything that happens after you find listings.
The touring. The note-taking. The landlord outreach. The cost calculation. The comparison. The decision. That entire workflow — which takes 20 to 40 hours for an average renter — lives in a graveyard of scattered tools: browser tabs, Google Sheets, phone cameras, email, text messages, and gut feelings.
And it affects a lot of people. There are 44 million renter households in the United States. Twenty million of those are people under 35 — students, new grads, young professionals — who move more frequently, have less experience, and are more vulnerable to bad decisions and outright scams. The FTC reported $65 million lost to rental scams last year. Renters pay $3.7 billion annually in hidden junk fees.
Yet every proptech AI tool on the market serves landlords, not renters. The landlord gets AI-powered pricing optimization, automated tenant screening, and smart leasing workflows. The renter gets a Google Sheet.
So I Built the Tool I Wished We’d Had
Snugd is an AI-powered apartment hunting platform for students and young professionals. It replaces the scattered workflow with a single pipeline that tracks every apartment from first discovery through your final decision.
Here’s what it does. When you search, AI scores every listing against your priorities — budget, commute, amenities, neighborhood — and gives each apartment a match score from 0 to 100 percent. When you find apartments you like, you heart them into favorites and move them into a five-stage touring pipeline: Interested, Outreach, Scheduled, Toured, and Deciding.
For outreach, AI drafts your inquiry message. You choose how to send it — phone call, text, or paste into the listing’s contact form. The message auto-trims to fit each channel’s character limits. During tours, you hold a button and talk. Whisper transcribes your voice notes and AI organizes them into structured pros and cons. You snap photos and they’re linked to each apartment automatically.
After touring, the True Cost Calculator shows you what you’ll actually pay — rent plus utilities, insurance, and fees, with per-person splitting if you have roommates. Regional estimates fill in what the listing doesn’t tell you. The Commute Calculator shows drive, transit, and walking times from each apartment to your work or school.
When you’re ready to decide, you pick two or three finalists for a side-by-side AI comparison. Claude analyzes each across value, space, amenities, and location, then picks a winner with detailed reasoning. And if you want the full picture, a Decision Brief synthesizes all your ratings, notes, tags, and tour data into a recommendation.
What I Learned Building It
Three lessons that might be useful if you’re a founder building an AI product.
First, the hardest part isn’t the AI — it’s replacing the workflow. Claude can score an apartment in under 60 seconds. But getting a renter to trust an AI recommendation over their gut feeling? That takes a pipeline they can see, edit, and control. The touring stages, the voice notes, the tags and ratings — those exist so the user feels ownership of the data the AI uses to make its recommendation. AI that works with you beats AI that works for you.
Second, voice-first changes everything for mobile. Nobody wants to type notes while walking through an apartment. When we added hold-to-record with Whisper transcription, engagement jumped. Talking is natural. Typing is work. If your product involves capturing information in a physical space, try voice.
Third, start with the wedge that has a deadline. New grads relocating for rotation programs need apartments by a specific date. That urgency compresses the decision window to two to four weeks. If your product reaches them before that window opens, you have their full attention. If you miss it, you wait a year. We’re targeting employer rotation programs — PNC, Deloitte, the Big Four — and distributing invite codes through their HR teams. Every cohort is a built-in batch of users who all need the same thing at the same time.
What’s Next
We’re in beta across 15 cities — including Boston, which is one of our most active markets. The business model is freemium SaaS: free tier for basic search and tour tracking, Pro at $12 a month for the full AI experience. Future revenue comes from affiliate partnerships with renter’s insurance and utility providers at the point of lease signing — a conversion moment no other platform owns.
I’m also a co-founder of GirlHacks, an organization empowering women in STEM through hackathons and mentorship programs. Building community-driven products isn’t new to me. Snugd started the same way GirlHacks did — by watching someone I care about struggle with something that should be easier, and deciding to build the fix.
If you’re apartment hunting in Boston right now — or know someone who is — I’d love your feedback. Visit snugd.ai to try it.
And if you’ve ever apartment-hunted with a Google Sheet, you know exactly why this exists.
About the author: Shuchi Agarwal is the founder of Snugd (snugd.ai) and co-founder of GirlHacks (girlhacks.net). She’s a software engineer, a UPitt parent, and a recovering Google Sheets apartment tracker.
