One is a one-person practice for now — by design. Bryce Galen has spent his career at the intersection of consumer software, fintech, and real estate, with stints in games and B2B tech along the way. Below: the long version.
I run Landian, the first flat-fee, AI-driven realtor service for buying, selling, and renting homes across the US and Canada. Today: 850+ partnered real-estate agents across 64 jurisdictions, and a single flat fee, often $4,999, in place of the standard 3.0% agent commission — on a million-dollar home, that's $25,000 back in your pocket. It helps buyers, sellers, and renters keep more of their own money: buyers get cash back at closing, sellers save tens of thousands on commission, and renters pay less in fees.
One is the consulting practice that grew out of operator friends asking me to look at their products. It's selective — a handful of engagements at a time — and most of the work lives at the intersection of product strategy and AI feature engineering.
Before Landian, I founded Zero — the consumer fintech that pioneered the “debit-style credit card.” It paid customers up to 3% cashback while keeping the overspending guardrails of debit. Zero won a Real Simple Smart Money Award in 2020, raised $34M, grew to 50,000 customers, and was acquired by Avant in 2021.
At Zynga I was a Senior PM on Words With Friends, where I led competitive leaderboards, the community matchmaking system, and the launch of “New Words With Friends” — the franchise refresh that pushed the game back to the top of the App Store charts and reset its trajectory for the back half of the 2010s.
Earlier, I founded TriviaTryst, a NYC-based mobile trivia game that went through the consumer-startup gauntlet of launch, retention work, and acquisition. Before that I was in real-estate private equity at Fidelity Investments, doing diligence and investments across a $1B value-add fund spanning every major property type, and started my career in investment banking at Wachovia Securities.
I started shipping AI features in 2017 — long before the ChatGPT moment made it a hiring keyword. The lesson from those years: most AI features fail not because the model is bad, but because the product around the model is bad. The interesting work happens at the edges: evals, retrieval, prompt versioning, agent loops, the small-model handoff, the failure-mode design.
In 2026 the table-stakes have moved. “We have AI” is no longer a pitch — it's an assumption. The teams pulling ahead have taste, discipline, and patience: taste to pick the right use cases, discipline to ship evals before features, and patience to design for failure modes that only show up in production. That's the work.
I read every email. No bots, no auto-responders.
hello@talktoone.com →