12·Applied AI
Barnes Dubai LLM
A domain-tuned LLM and agentic broker assistant for Dubai luxury real estate, a QLoRA fine-tune on 15,369 instruction pairs, running locally.
Ran in the Dev Portal playground until BARNES Dubai ceased operations in 2026.
[ Ask about this build ]Why it exists.
Data Centre surfaces the data; Barnes Dubai LLM reasons over it for the broker asking the question. Fine-tuned on real-estate jurisprudence, deployed locally via Ollama at zero per-query cost, reachable through WhatsApp or the Dev Portal's in-portal Barnes Dubai LLM assistant.
Brokerage-grade output needs DLD transaction conventions, France DVF reporting cadence, and Singapore URA mapping baked in, none of which live in a frontier model's pre-training in a useful way. The instruction corpus taught the base model what BARNES Dubai means by recent, comparable, or flag for review.
The brief.
Decide whether a domain fine-tune beats base + RAG for a terminology-heavy, mixed-language vertical, then ship the answer brokers can use on WhatsApp.
What it took
Skills behind it.
Primary discipline plus the support stack.
Skills demonstrated
- AI & Machine Learning
QLoRA domain fine-tune plus a tool-calling loop (searchListings, searchBarnesListings, searchOpportunities, getOpportunityDetail, draftWhatsAppMessage, among others) grounded by a system prompt that instructs the model never to answer from training data or invent a listing.
- Prompt Engineering
Prompt + tool-schema design for the agent loop, system prompts, function-calling contracts, and an eval harness over real broker queries.
- Backend Engineering
FastAPI inference service with broker_id-keyed conversation memory; Ollama deployment; WhatsApp surface in front of the playground.
- Data Pipelines
15,369 instruction pairs from five jurisdictions' open data, with live Data Centre connectors and the BARNES listings database feeding tool calls.
- Research
Training-run analysis and dataset assembly that targeted brokerage-grade jurisprudence.
The build, in full.
How it came together and the decisions on the record - kept off this page so the case study stays a read, not a scroll.
Read the full build notesWhat shipped.
Ran in the Dev Portal playground until BARNES Dubai ceased operations in 2026. A broker asked Compare Business Bay vs JVC for buy-to-let yield and got a tool-resolved verdict in seconds (medians, gross yields, tenant pool, a one-paragraph recommendation), the brief that used to take an analyst an afternoon.
Also reached brokers via WhatsApp: the same agents behind a number a broker already messaged. Under every answer sat three national property registries (Dubai DLD, UK Land Registry, France DVF) and the BARNES listings database, resolved through tool calls rather than recall.
QLoRA fine-tune corpus
Dubai DLD, UK Land Registry, France DVF
The paper-to-product thread this project sits inside. The full chain, paper through benchmark to shipped, is on the thread.
- Domain LLM fine-tuneSee thread
Teaching a general-purpose model the vocabulary of one market, on one GPU.
A look inside.
01AI Assistant
See also
Data Centre
BARNES Dubai's internal data tooling: an entity-centric valuation platform and AVM, absorbed into a broker tool over ~840,000 ownership records.
Footing
A white-label AI co-pilot for US luxury brokerages: one plain-language interface over the MLS, CRM and public records. Code complete, backend offline.
Barnes Dubai Dev Portal
The internal cockpit BARNES Dubai's innovation programme ran on: project tracking, analytics and a marketing-engineering toolkit in one place.