Larracos Labs
Building data and AI-driven companies in complex problem spaces
Larracos Labs is a product-led venture studio. We build software in domains where decisions matter, data is abundant, and existing tools fail to connect the two. Our products turn messy, real-world data into structured intelligence that people can actually trust.
Explore our venturesVentures
Current ventures
LiveThe right restaurant, every time
Seemor helps people pick a restaurant with confidence — for the occasion, for their taste, in seconds. It analyses every restaurant across 35+ dimensions of quality, then personalises results so your recommendations reflect what you actually care about.
Two ways in: describe what you're planning in plain English and get a confident recommendation, or explore the map with filters you won't find anywhere else: noise level, service speed, kid-friendliness, authenticity, and more.
The difference is underneath. Most tools rank restaurants by aggregating other people's reviews. Seemor analyses each restaurant itself, building a structured profile across every dimension that shapes a night out, and sharpens that profile as more people use it. That data asset is what lets it recommend with confidence instead of handing you a list and leaving the judgement to you.
250,000+ analysed restaurants · hundreds of cities · 60+ countries
Project Atlas
Atlas is an analytical tool that makes the structure of disagreement visible. On any contested issue, it shows what is broadly agreed, what is disputed, where uncertainty is highest, and whether people are disagreeing about facts, interpretation, values, or confidence thresholds.
It does not issue verdicts or score bias. It decomposes issues into atomic assertions and classifies why groups reach different conclusions, giving people the clarity to disagree accurately.
Explore the prototype →Sorted
Every family with school-age kids drowns in email. Eighty-plus messages a month, each burying a date, a permission slip, or a deadline somewhere in the middle. Miss one and it's the forgotten costume on dress-up day.
Sorted reads the flood for you. Forward the emails you care about and it pulls out what matters: dates land in your calendar, tasks land in a list with deadlines, and a short morning briefing tells you what is coming.
It only ever sees what you forward. Not your bank, not your work, not the rest of your inbox. Handing an AI agent your entire mailbox is a bet most people should not have to make, so we built one that never asks you to.
We started with school because that is where the pain is sharpest, but the same approach fits any firehose of email you would rather not read in full. It is the Larracos pattern in the most relatable place we could find it: messy real-world data in, structured intelligence out, a better decision at the other end. The same thing Seemor does for where to eat, pointed at the chaos of family logistics.
What we work on next
We are always looking for the next problem worth building around — domains where decisions are hard, data is abundant but poorly understood, and AI can create a step-change in how people navigate complexity.
If you have deep domain expertise and see a problem that fits this pattern, we'd like to hear from you — whether you're an experienced founder, an operator who's been thinking about going out on your own, or someone who just sees something broken that nobody is fixing well.
Larracos provides the data, AI, and product infrastructure. The right person brings the domain insight and the drive to build.
How we build
- •Real problems first. We start with genuine pain and unmet needs, not novelty.
- •Data as leverage. We structure, model, and apply data to create lasting product advantage.
- •Small, senior teams. We move quickly, validate early, and iterate with discipline.
- •Practical AI. We use models where they add durable value, not decoration.
- •End-to-end ownership. From zero to product to scale, with spin-outs when it makes sense.
Capabilities: Product strategy · Full-stack development · Data and AI systems · Early GTM experimentation · Growth foundations
About
Ryan Fuller—Founder
Ryan Fuller is a founder and product builder with over 25 years of experience creating data-driven software and leading teams at scale.
He began his career as a software engineer working on data and analytics systems, then went on to co-found and scale venture-backed startups. His first company, VoloMetrix, was acquired by Microsoft, where he later served as a Corporate Vice President. At Microsoft, Ryan built and led data- and AI-driven products from zero to over $100M in annual revenue and led a 300-person global organization, helping establish the data and intelligence foundations for Microsoft's modern AI platform.
Ryan founded Larracos Labs as a vehicle for building new data- and AI-driven companies and for selectively working with founders and leadership teams at moments where judgment, synthesis, and execution matter most.
Advisory work
In addition to building companies through Larracos Labs, Ryan works selectively with investors and leadership teams at moments of transition or inflection, where clarity and judgment materially change outcomes.
His advisory work focuses on helping organisations clarify what actually matters when signals conflict, data is messy, and execution is stalling. Typical engagements span product and technology strategy, data and AI foundations, organisational alignment, and early go-to-market decisions.
Ryan is most useful where judgment, synthesis, and practical operating experience are needed, rather than functional optimisation or generic advisory support.
If you believe that perspective could be helpful, feel free to get in touch.