Data work at RentLondonFlat.com: from app schema to dashboards
Almost three years looking after the company's data, from designing the tables its in-house app writes to, through to the dashboards the team reads. Along the way: checks that keep bad records out, a health score for every property, a value score for every customer, and a GPT that turns a plain-English question into a query.
From February 2023 to December 2025 I worked at RentLondonFlat.com as Data Architect and BI. The role covered the company's data end to end: the application database, the pipelines into the warehouse, data quality, business metrics and dashboards.
Application schema
The company's in-house internal application needed backend tables and a schema. I created both.
The app was still in its design phase, so the requirements kept changing, and the schema had to keep up with them.
The SQL procedures I wrote propagated an action performed on one table automatically into the related tables. That kept the tables consistent without anyone updating them by hand.
Pipelines
Data moved from the Postgres backend into BigQuery through incremental pipelines I built. Each run loaded only new and changed records instead of reloading everything.
Data quality
Ingested data went through cleaning and quality checks: validation of names, emails and postcodes, duplicate removal, and anomaly detection. The checks ran at ingestion, so the dashboards and metrics downstream worked from clean data.
Metrics and dashboards
The property health metric covered every property the company listed. It answered one question: is this property in order? Were the required certificates and forms in place, when did each one expire, and how much time was left to renew it. A property with a document close to expiry showed up before the deadline, so the team could renew it or inform the owner in time.
The customer value metric did the same for customers. It combined signals such as their job, location, the device they used, paying capacity, past experience with the company, repeat-customer status and potential revenue into one score.
I created both metrics. On the reporting side, the work was Looker Studio visualizations for the company dashboards and BI dashboards in Looker.
SQL generation with a custom GPT
The starting point was a schema mapping of the relationships between the company's tables. On top of it I built a custom GPT for the company that used the mapping to generate SQL from questions asked in plain language.
Tools
Postgres, BigQuery, SQL, Looker, Looker Studio, custom GPT.