Starting with AI-native listings, Alma expands layer by layer across discovery, the taste graph, tours, diligence, offer, settlement, and lifelong home memory.
It is the system that helps buyers and sellers discover, evaluate, tour, negotiate, close, and remember every home.
Alma starts with the listing because that is where the transaction first becomes real: seller knowledge, home history, and buyer intent finally start accumulating in one place.
The first product gives every listing a beautiful home website and a 24/7 AI that knows the home. From there, Alma expands layer by layer until buying and selling feels like one coordinated system.
Portals know search behavior. Agents know conversations. Lenders know financing. Inspectors know risk. Title and escrow know closing mechanics. Sellers know the home itself. But none of that context becomes a shared operating layer for the transaction.
The buyer rebuilds preferences on every portal. The seller compresses years of lived knowledge into a few MLS fields. The agent becomes the manual router for answers that should already be available. Diligence starts after the offer instead of shaping the offer. Title, insurance, financing, inspection, and closing all arrive downstream from the decision, when the buyer is already emotionally and financially committed.
The MLS problem is the first symptom: homes are still treated like rows in a database. But the deeper problem is that the transaction has no memory.
Alma starts where the buyer first meets the home because that is where transaction memory should begin.
Alma owns the full stack the transaction touches.
| Layer | What Alma owns | What changes |
|---|---|---|
| Discovery | The agentic matching layer between buyers and Alma's listing network. | Buyers stop searching portal rows. They describe what they want in human terms, and Alma surfaces the homes that fit, why they fit, and which are worth digging into. |
| Taste graph | A per-buyer preference model built from real conversation across homes. | Buyers are understood by what they ask, compare, and care about, not by filters. |
| Tours | The agentic access layer for in-person, remote, and self-guided showings. | Alma qualifies intent, coordinates access, guides the buyer through what to notice, and feeds tour signal back into the buyer's taste graph. |
| Diligence | The pre-offer trust layer: a listing-time data room for disclosures, inspection, title, HOA, insurance, systems, renovations, seller knowledge, and verified property records. | Buyers get the hard questions answered before they offer. Sellers get fewer retrades, fewer surprises, and cleaner offers from buyers who have already seen the record. |
| Offer and settlement | A trusted settlement layer with verified buyer funds, listing-time diligence, programmable escrow, and ownership transfer in one flow. | Buyers make offers after the data room, sellers can trust funds and terms up front, and the happy path moves from accepted offer to settlement in minutes instead of weeks. |
| Lifelong companion | An agent for living in, maintaining, improving, and eventually selling the home. | Alma remembers systems, maintenance, upgrades, receipts, costs, renovation plans, and the lived story of the home. When the owner is ready to sell, the listing is already there. |
Alma gives every listing a home website that feels bespoke and ships in minutes.
An agent uploads the listing. Alma turns photos, documents, MLS data, and seller knowledge into a dedicated website with a 24/7 AI assistant trained on both the facts of the home and the feeling of living there. Buyers can ask about the roof, renovations, utilities, schools, flood risk, and maintenance, then keep going: morning light, quiet corners, dinner parties, storage tradeoffs, the way the kitchen works during a busy week, and the little details that make a home either right or wrong for a life.
The assistant helps buyers understand the home before it asks them to become a lead. Serious buyers arrive warmer because they have already asked the questions that matter.
Agents get listings that stand out and better-informed leads. Sellers get a first impression that carries the home's story. Buyers get answers at 10pm instead of waiting for a callback.
The page is not the moat. The wedge starts the transaction graph portals do not have: seller knowledge, property records, diligence artifacts, agent context, and buyer questions tied to specific homes before the offer.
Phase zero was making the AI-native listing work. That is done. The seed round funds phase one: build enough listing density in one market to unlock buyer pull, taste graph signal, and the next transaction layer.
Alma is going deep in a small number of markets where we have a reason to win. Orlando is the first density test because Jessica and the association path are already active. Atlanta and Bozeman are also in scope because they connect to founder and GTM relationships.
In parallel, Alma is pursuing Realtor association partnerships that can onboard many agents through one relationship. The goal is not broad national spray. It is finding the markets where an association, a champion, or a local partner can turn agent adoption into listing density quickly.
If partnerships land, they accelerate the market plan. If they do not, Alma can still build through direct agent GTM in a small number of focused markets. The seed round keeps both tracks in play for 18 to 24 months.
Brian spent twelve years at Ceros as CTO/COO, helping take the company from zero to $50M ARR, a 100-person product org, and a PE exit. He owned product and engineering, and operated across strategic planning, GTM, customer success, business systems, and AI transformation.
Alma came from living the broken transaction. Brian sold his Atlanta home, bought in Puerto Rico, and hit every broken layer directly. The yard sign sent buyers to a spammy text-to-short-code flow that broke its promise before they even saw the home. Simple buyer questions became a four-hop telephone game. Listing pages could not explain what made a home special. Diligence arrived late, scattered, and stressful, when it should have been alive from the first serious buyer interaction.
Pre-money SAFE. Phase zero is done: the AI-native listing works. The seed round funds phase one: turn the live wedge into a listing network.
Over 18 to 24 months, Alma will run two GTM tracks in parallel: focused market density and Realtor association partnerships.
The Series A bar is one focused market where listing density produces repeat agent usage, measurable buyer engagement, taste graph signal, and the first diligence workflow attached to live transactions.
Market density, association partnerships, agent onboarding, Orlando execution, and the first repeatable playbook for Atlanta, Bozeman, or the next partner-led market.
Product depth, onboarding, AI quality, listing network infrastructure, data room foundations, and early trust and settlement primitives.
This is a small team by design: a GTM lead with deep partnership experience is ready to step in as soon as the round closes. An engineering lead scales the platform while Brian stays close to product, company, and capital.
Residential real estate moves more than $2T a year in the United States and leaks roughly 10% of transaction value across agents, portals, lenders, title, escrow, insurance, inspectors, attorneys, and coordination work.
Alma's end state is a 2% transaction: 1% per side, with discovery, diligence, access, settlement, and ownership memory running through one system.
The wedge gets Alma onto the listing. Density turns listings into a buyer network. The data room and settlement layer let Alma participate in the transaction itself.
After close, the lifelong companion gives Alma a durable consumer relationship through subscription revenue and the spend that already flows through the home: maintenance, repairs, upgrades, renovation planning, receipts, system history, and the next sale.
The prize is transaction revenue during the sale, subscription revenue during ownership, and the operating layer for the home before, during, and after the transaction.