Dominic Mayne
Real Estate Graduate & Proptech Builder
BSc Real Estate, University of Reading (2:1). I build data products that put real estate knowledge to work, from a Central London office intelligence platform to a live service used by parents across the UK.
Why I built this
A real estate degree, a gap in the market, and a question about what AI could do with it.
I graduated from the University of Reading with a BSc in Real Estate (2:1), with modules covering Real Estate Valuation, Development Appraisal, Applied Property Law and Asset Management. My final project received a First Class grade.
During my studies and job search I noticed a gap: the tools used by institutional property professionals — CoStar, EG Propertylink, agent research platforms — are expensive, complex, and largely inaccessible to graduates. I wanted to understand how modern technology could democratise access to this kind of intelligence, so I built my own version.
I had no coding background, so I built it the way I think a lot of work will be done from now on: I researched the data, wrote the specs, made every product and design call, and directed Claude to write the code. That meant learning enough Python, FastAPI and scikit-learn to read what it wrote, test it, and push back when it was wrong.
It worked well enough that I kept going. Hold now helps parents across the UK get their children into swimming lessons, and Unearthed turns scouting young footballers into a game.
Alongside the building, I've kept close to the industry: work experience at Savills, Fletcher Morgan and PMG in Cardiff, and in June 2026 at Union Street Partners in London, where I calculated business rates for disposals, reviewed Heads of Terms and joined office and industrial viewings across the Southbank.
I'm now looking to bring both sides of this — real estate knowledge and technical capability — into a graduate role at a firm where data, technology and property intersect.
How the platform works
A full-stack proptech platform: data pipeline, a backtested forecast, AI integration and a frontend, all deployed.
Data Research & Pipeline
Researched and compiled 11 quarters of market data (2023 Q1 to 2025 Q3) across 8 Central London submarkets from published reports by Savills, CBRE, Colliers, Knight Frank and BNP Paribas. Built a structured CSV pipeline feeding a Python backend.
PythonPandasMarket Research
Machine Learning Forecasting
A ridge regression forecasts next quarter's vacancy change from momentum and relative take-up, four quarters ahead. Walk-forward backtesting shows it beats a no-change guess by 19%.
scikit-learnRidge RegressionBacktesting
FastAPI Backend
Built a production REST API with FastAPI, deployed on Railway. Endpoints serve forecasts, live web data, deal tracking and PDF generation, with web-search responses cached to keep costs down.
FastAPIRailwayREST API
Live Market Search
The Anthropic API with web search pulls the latest published vacancy and rent figures and recent deal transactions, so the recorded series can be checked against what agencies are reporting now.
Anthropic APIWeb SearchPrompt Engineering
Frontend & Product Design
Vanilla HTML, CSS and JavaScript: an interactive submarket heat map, Chart.js visualisations, light and dark themes, and a saved data snapshot so the pages never sit blank when the API is down.
HTML/CSS/JSChart.jsSVGVercel
PDF Report Generation
Built an automated PDF report generator using ReportLab that produces a branded, institutional-quality research note on demand, with a written summary built from the data, KPI tables, the quarterly breakdown and a disclaimer footer.
ReportLabPDF GenerationDocument Design
Everything else I've built
Different problems, same approach: find a real question, get the data, ship something people can use.
Kids' class alerts for parents · Founder
Founded in 2026. Watches swimming-lesson timetables at Better (GLL) and Everyone Active leisure centres across the UK, checking every 60 to 90 seconds, and emails parents the moment a space opens. First users came through a Google Ads campaign; now growing into other activities. gethold.co.uk ↗
PythonPlaywrightRailwayEmail automation
Fantasy football for wonderkids
Pick seven real under-21s and score from what they actually do each week, in mini-leagues with friends. A career mode simulates five teenagers' next ten seasons.
Node.jsAPI-FootballGame design
Built Environment Talent Intelligence
Salary benchmarking and hiring-demand forecasting across five property, engineering and construction recruitment sectors. View the demo
PythonChart.js
Stadium Commercial Intelligence
The property and revenue model behind Tottenham Hotspur Stadium: NFL, concerts, boxing, F1 Drive and a £1.2bn real estate case study, with five years of revenue from the club's annual reports and the Deloitte Football Money League, benchmarked against Manchester City, Arsenal and Liverpool. View the dashboard
Stadium economicsChart.js
Check-in and check-out, in 3D
For UK lettings: an agent films a slow walk round each room on a phone, Samespot builds a 3D map of it, and at check-out it finds what changed and shows each mark beside the check-in photo of the same spot. Every photo is fingerprinted, so the dispute pack sent to myDeposits, TDS or DPS can be shown to be untouched. Working prototype; agent portal built.
3D reconstructionChange detectionClaudePropTech
Judgement in, software out.
-
Start with a real question
Every project began with something I noticed. Graduates can't afford CoStar. Parents can't get a swimming place. Fans love spotting talent before anyone else.
-
Direct the build
I write the spec, make the product calls and direct Claude to write the code. I had no coding background, so I learned to read, test and question everything it produces.
-
Ship it and be honest about it
Everything here is deployed with real data. The West End forecast is backtested against a simple baseline, and Hold publishes how it measures what it reports.
Technical & professional capability
Built through the project and underpinned by a real estate education.
Education & work history
Work Experience
- Calculated current business rates for disposals using the required formula, so the figures marketed were accurate; a wrong figure risks a breach of the Misrepresentation Act 1967
- Joined office and industrial viewings with prospective tenants across the Southbank and SE1, from best-in-class Cat A offices to warehouses on short leases
- Built a Competing Buildings document from scratch for a client presentation, researching the metrics for each alternative and presenting them in PowerPoint
- Reviewed Heads of Terms for an office letting: parties, transaction, price and payment, key warranties, conditions precedent, and the binding exclusivity, costs and confidentiality clauses
- Represented USP London and Tuckerman at The Commercial Property Network, meeting people from firms across the UK
Chartered Surveying Experience
- Shadowed a company Director on multiple site visits with current and prospective clients
- Supported the Director with administrative tasks on site visits
- Saw the practical, day-to-day work of a Chartered Surveyor and a company Director
Residential Property Insight
- Met the Directors of a founder-run residential business specialising in high-quality rental accommodation, for advice on the industry
- Saw how a small, family-run property firm operates, and how that differs from larger firms
BSc Real Estate — 2:1
- Modules included Real Estate Valuation, Applied Property Law, Development Appraisal and Finance, and Project in Valuation and Asset Management
- Final year project received a First Class grade
- Regular group presentations and public speaking throughout the degree
Work Experience
- Researched and analysed an upcoming multi-million-pound project, supervised by colleagues
- Observed a meeting with the project's developers, seeing how firms take on and run schemes of this kind
- Shadowed professionals on site visits, watching how they worked with potential clients
A-Levels: Economics, History, English Literature
- Awarded an Academic Scholarship