Selected Work

AI systems, automation, and software we've built.

Explore the business problem, what we delivered, and how each system works. Filter by industry to find projects relevant to your team.

3 projects in Real Estate & Property

Real Estate & Property

Document-heavy operations where analyst hours and lease turnaround decide the margin.

Real Estate · Document AI

Real Estate Document & CRM Intelligence

The Challenge & What We Built

A real estate investment firm was losing expensive analyst hours to document handling and manual CRM entry. We scoped a phased automation programme sequenced so value arrives in week one rather than month three: an immediate flyer-generation win, then CRM auto-population through a property data API, then an AI review pipeline for Letters of Intent and Purchase & Sale Agreements that extracts and checks terms against the firm's own positions, and automated triage of inbound listing email.

What Was Delivered

Analyst hours back on deals

LOI / PSA document AI

CRM auto-population

Value in week one

Built With

LLM Agents
Locate Alpha API
Document Extraction
Gmail API
n8n

Document Automation · PropTech

Automated Lease Generation Pipeline

The Challenge & What We Built

Approved rental applications arrive by schedule or webhook, data is fetched from Buildium and validated, leases are generated as HTML, converted to PDF, filed to Drive, logged in Sheets, and emailed to the leasing team.

What Was Delivered

Fully automated

Zero manual entry

End-to-end error handling

Built With

n8n
Buildium API
JavaScript
Google Drive
HTML-to-PDF

Sales Intelligence · Multi-Agent AI

AI Sales Call Auditing & Coaching Engine

The Challenge & What We Built

Managers at a student housing operator were reviewing sales calls by hand, which in practice meant most calls were never reviewed and the few that were got judged inconsistently. Recordings are collected continuously, transcribed, and filtered before two parallel scoring tracks run: an agent chain assessing opening, requirement discovery and objection handling, alongside an engagement agent rating rapport with verbatim evidence. An aggregation stage resolves both into a single comparable score, written back per caller into the CRM.

What Was Delivered

Every call held to one standard

Coaching from evidence

Comparable caller benchmarks

CRM write-back

Built With

OpenAI GPT-4o
gpt-4o-transcribe
Python
Google Drive API
CRM API

Have a similar challenge?

Tell us what's slowing your team down or what you want to build. We'll help you work out the next step.

Get a Free Audit Request a Discovery Call