Plot Analyzer
Message it five numbers about a raw plot - get back the feasibility study a consultancy would take weeks to assemble.
Design + Build - strategy, AI orchestration, engineering · Conversational AI agent · workflow orchestration
A feasibility engine for raw land, shaped as a consultant you can message. You hand it almost nothing - a point on the map and a few constraints - and it works the plot the way a senior advisor would: figure out where it really is, what the law lets it become, who lives around it, what already competes with it, and whether the numbers hold.
The whole engagement starts from five inputs: coordinates, plot area, GFA, the permitted number of floors, and a land price. No documents, no uploads, no site visit. Everything else, the engine goes and finds.
One line carries the whole idea:
The tools own the facts. The consultant owns the judgment. Every number is fetched and verified by a real data tool - the engine reasons over them like an advisor, but never invents one.
A consultant you message, not a form you submit
Underneath the chat is an analyst with a deliberate split personality: internally it reasons like a senior MENA investment consultant whose job is to protect your capital with honest, critical due diligence; externally it talks like an approachable assistant, in whatever language you wrote to it.
And it doesn't bury you. The lazy version of this dumps a hundred-page PDF on the first message. This one does the opposite: it runs only the fast checks first - resolving your coordinates into a real address, pulling current market context - hands back a quick expert read, then offers a numbered menu of deeper analyses and waits. You pick what's worth the wait; it runs that one specialist, reports back, and offers the next step. The depth is on demand, not on dump.
What it can dig into
Each track is a focused specialist the consultant can call, every one grounded in live data rather than guessed:
- Location & infrastructure - address, transport and transit, schools, healthcare, amenities, and zoning, rolled into an investment score.
- Competition - the competing and comparable projects nearby, their pricing, and where a new project would sit among them.
- Financial model - construction cost, revenue, net profit, ROI and IRR, run across best / base / worst-case scenarios off your land price.
- Product development - the optimal concept for the plot: target segment, unit mix, amenities, and positioning.
- Marketing - target audience, channels, budget, and go-to-market positioning.
- Research - live market data pulled on demand to ground any of the above.
A summary in the chat, the full study on demand
Every analysis comes back twice. In the chat you get a tight executive summary - the numbers and the decision, nothing decorative. Ask for the file, and the complete long-form report lands in your Telegram: the financial model alone runs to 25+ pages. Nothing is lost in the compression - the full study is always one word away, and stacking the tracks builds a complete feasibility dossier on the plot.
Honesty is the invariant
What makes it a tool a developer trusts is what it refuses to do. It never invents a price, a project, a developer, a distance, or a zoning rule - if a number isn't verified by a data or research tool, it doesn't appear. The senior-consultant instinct points the same way: surface the risks plainly, protect the capital, give the ground to decide. The judgment is the engine's job; the decision to build stays with the person.
Result
A private, now-archived engagement. The deliverable was the engine itself: a conversational analyst that turns five plain inputs into developer-grade, multi-page feasibility reporting on demand - the kind of cold, unfamiliar problem, entered with thin input and carried through to something you can actually decide on, that the studio exists for.
Private engagement - no public link.