// machine access

Give your systems and your AI direct access to property data.

Resolve an address, query a market, or feed your models, live, from the same data behind this site. MCP, API, and data feeds.

Get access →

resolve & enrich

# ask for offices in a precinct, with vacancy
GET /v1/properties?precinct="Sandton"&category="office"

{
  "precinct": "Sandton",
  "properties": 1284,
  "gla_m2": 3920000,
  "vacancy_m2": 512000,
  "as_at": "2026-06-01"
}

// illustrative response · property & aggregate data only, no personal or contact data (POPIA-safe)

Three ways to plug in.

Same data, whichever fits how you work.

MCP

MCP (Model Context Protocol) is the standard plug that lets an AI assistant such as Claude or ChatGPT use an outside data source directly. Connect yours to Gmaven, and it answers property questions from fact, not guesswork.

Best for: you and your team, working inside an AI assistant.

for AI assistants · conversational

API

Resolve and enrich a single address, or query a whole market, in real time. REST, JSON, keyed.

Best for: your systems, needing answers record by record, in real time.

real-time · per-call

CSV & data feeds

The full dataset, or the slice you need, delivered on a schedule and kept fresh.

Best for: your analysts, wanting the whole slice on a schedule.

batch · scheduled

What it looks like in use.

Three real jobs, one per channel. All values illustrative.

what you send Gmaven what you get back

MCP

Messy list in, clean data out.

Drop a spreadsheet of property names, coordinates or addresses into your AI assistant. Get back clean, enriched data: a unique identifier, the resolved property name, the street address, suburb, precinct, city, province, size and category, for every row.

Your spreadsheet

the meridian sandton?

-26.1076, 28.0567

135 west st sandown

ACME HOUSE

Gmavenvia your AI

Back to you, enriched

ID	Name	Suburb	Province	GLA m²	Category
GM-084213	The Meridian	Sandown	Gauteng	38 400	Office
GM-102467	40 On West	Sandown	Gauteng	6 250	Office
GM-077310	135 West Street	Sandown	Gauteng	12 800	Office
GM-118904	Acme House	Sandown	Gauteng	4 100	Office

+ street address · precinct · city · names and figures illustrative

API

One question in, one answer out, instantly.

Send a coordinate pair from inside your own system. Get back the suburb name, the average asking rental for that grade, the vacancy rate for that grade in that precinct, and whether rentals are rising or falling.

lat -26.1076 · long 28.0567

Gmavenlive

Response

suburb "Sandown"

precinct "Sandton"

avg_asking_rental_a_grade R230 / m²

vacancy_a_grade_precinct 9.4%

rental_trend ▲ rising

Batch

The whole slice, on schedule.

Receive the latest solar penetration, aggregated by suburb, city and province: by lettable area, by installed capacity (kilowatt-peak), by property count, as a percentage of all properties, with available roof space. Delivered on your schedule, kept fresh.

your schedule

solar_by_suburb.csv

solar_by_city.csv

solar_by_province.csv

In each file: lettable area (GLA) installed capacity (kWp) property count % of all properties available roof space

Every answer resolves to one property.

Behind all three channels sits the same mastered database: more than 64,000 commercial properties, each resolved to one persistent, unique identifier, with a name, an address and an as-at date. Whichever channel you choose, you are reading the same single source. Choosing a channel is a workflow decision, not a quality trade-off.

Property and aggregate data only.

What travels over these channels is property and aggregate market data: no personal information.

Get access.

Tell us what you want to plug in, and how you work. We will recommend the channel and scope it with you.

Get access →

Looking for space rather than data? See Find space.