houseraicoreNumbers your board can question and your team can defend
We work with landlords, housing associations and asset managers on the unglamorous half of analytics: getting records straight, agreeing what a metric means, and keeping models honest after launch.
Illustrative pattern from anonymised engagements, not a live feed.
Three things clients tell us matter
We learn the estate first
Before any modelling, we sit with the people running repairs, lettings and compliance. Most data problems turn out to be process problems wearing a disguise.
Every figure has a lineage
If a number appears in a board pack, you can trace it to the source record and the rule that transformed it. Auditors ask; we would rather you had the answer ready.
We plan our way out
Contracts include handover: documentation, runbooks and paired work with your staff. A consultancy that cannot be removed is a liability, not a partner.
Four ways engagements usually start
Most clients begin with one and add others once they see how we work. There is no bundle you have to buy.
Data platform build
Ingestion from housing management systems, finance and IoT sensors into a warehouse with tests, documentation and a schedule people trust.
- dbt
- Snowflake / BigQuery
- Airflow
- Data contracts
Applied machine learning
Stock condition forecasting, arrears risk, repair demand. We ship the monitoring alongside the model, because a silent model is worse than none.
- Forecasting
- Drift monitoring
- Backtesting
- Model cards
Specialist staffing
Contract, contract-to-permanent and direct hire for data and platform roles. We only recruit in fields we deliver in, so screening is done by practitioners.
- Contract
- Contract to hire
- Direct hire
- Statement of work
Governance & audit readiness
Retention rules, access control, lineage and the paperwork a regulator asks for. Written so the teams affected can actually follow it.
- Lineage
- Retention
- Access reviews
- Regulator packs
What a typical year looks like
What manual reporting costs you a year
Rough arithmetic, not a quote. Move the sliders to your own numbers and the maths stays visible below.
Salary plus employer costs, divided by working hours.
Annual staff cost of assembling reports by hand
Method: hours × 52 × rate. The saving range applies a 55–75% reduction, which is the spread we have seen once pipelines replace spreadsheet assembly. Your figure depends on how much of the work is judgement rather than copying. Treat this as a conversation starter.
In their words
They spent the first fortnight in our repairs office rather than presenting slides. The model that came out of it reflected how the team actually triages jobs.
We had three dashboards giving three answers for void loss. houseraicore made us agree a definition, then rebuilt the pipeline behind it. Dull work, and exactly what we needed.
The two contractors they placed knew the sector. Neither needed a month of context before being useful, which is rare in my experience of agencies.
At handover we got runbooks and a fortnight of paired work. Six months on, my own team maintains it without calling them. That was the point.
Asked on most first calls
How long before we see anything working?
Do you replace our existing BI team?
Our data is a mess. Is that a problem?
Can you work with our housing management system?
How do you charge?
Where are you based?
Tell us what is not working
A thirty minute call is usually enough to tell whether we are the right people. If we are not, we will say so and suggest who might be.