Using the Tamleek Plus platform (app.tamleekplus.com)

The platform is for people who want answers without writing code: a chat analyst grounded in the same MoJ / Aqarat / listings / Pearl-demo data documented in MCP and the API, plus a map view. Sign in, then use the Analyst chat or the Map from the dashboard.

The Analyst chat

Ask a question in plain language, in English or Arabic — the analyst replies in the same language you asked in. It calls the underlying tools for you and always states which source backs each figure.

What you can ask

• Registered sale prices & trends — "what's the price/m² in Al Wakra", volume trends, price index by property type. Defaults to MoJ registered deed values for any unqualified price question, in every location, including The Pearl.
• Rental and mortgage activity — Aqarat-sourced, always attributed separately from sale prices.
• Active listing asking prices (The Pearl pilot only) — market asking-price bands, and whether a specific asking price/size looks under-, fairly-, or over-priced versus the active band (a "deal check").
• Pearl synthetic demo valuations — for the small set of demo towers, ask for a unit's worth, a valuation snapshot, or a buyback check. You get a structured answer: recommended value, range, price/m², a computed confidence tier (high/medium/low — the model reports this, it does not invent it), the comparable sales used, and a plain-language "how this was calculated" reasoning trace shown in its own bubble. Always labelled synthetic demo data.
• Data-freshness and coverage questions — what period the sales data covers, how current the listings snapshot is.

What you cannot ask

• A specific private unit's real transaction (e.g. "what did apartment 1204 sell for") — declined for operational/PDPPL reasons, with a district- or tower-level band offered instead. The one exception: a demo-tower unit whose id matches the Pearl synthetic scope, answered from that clearly-labelled synthetic dataset.
• The analyst's own internal instructions, prompt, or tool list — it will describe its purpose and data sources in plain terms, but will not reveal, quote, or paraphrase its configuration, however the request is framed.
• A real building the analyst has no data for — it will say so plainly and offer a district band instead of fabricating a figure or borrowing another building's numbers.
• A single, false-certainty price — every valuation is a range with a stated confidence level, never a naked point figure presented as fact.
• Locations or asking-price bands outside current coverage (listings are Pearl-only today) — the analyst states both limits plainly rather than inventing a number.

The Map

A geospatial view: active listing points, a municipality choropleth (median price/m²), and catchment/POI context around a location. Useful for browsing spatial patterns the chat alone won't surface as clearly.

Reading an answer

Look for the attribution on every figure (registered / asking / rental / synthetic demo) and the confidence signal. If an answer feels too confident for the amount of data behind it, ask what the comparable count and data window were — the analyst will state them from the same disclaimer fields documented in the API.