Portfolio — AI Automation & Operations engagement Bringing Order to a Property Management Operation That Had Outgrown Itself A Cape Town-based property management company handling around 100 properties with a team of five was drowning in manual coordination. Enquiries went unanswered, maintenance requests came in from every direction, and data lived in three disconnected tools. A 9-week engagement combining operational redesign and AI-powered workflow automation to replace the chaos with a system that ran itself.
Sector
Property Management
Market
South Africa
Stage
Owner-managed
Engagement type
AI Automation & Ops Build

The challenge

At around 100 properties, the business had reached the point where manual coordination was no longer sustainable. Enquiries from potential hosts were sitting unanswered for three or more days because follow-up ownership was unclear. Maintenance requests arrived simultaneously via WhatsApp, email, and phone, going to different team members with nothing tracked centrally, issues were being missed and hosts were chasing for updates. Tenant information was being manually copied between Gmail, a Google Sheet, and an invoicing tool, consuming hours every week. Leads that didn't convert immediately were forgotten entirely, with no follow-up workflow in place. The owner understood the business was losing time and likely revenue, but wasn't sure which tools made sense or how to connect them. They needed someone who could both diagnose the operation and build the solution, not hand over a slide deck.

Approach

Having built and run operations at Hostmaker across Lisbon and Barcelona, I recognised the pattern immediately. These are the exact failure modes that emerge when a property management operation scales past what informal coordination can hold together. I started with a two-week discovery phase: mapping how enquiries, maintenance requests, and host communication actually moved through the business day-to-day, where handoffs broke, and what the team's real bottlenecks were. From there I designed an AI automation and workflow automation layer on top of HubSpot CRM, using Zapier to connect the existing tool stack and deploying agentic workflows to handle the repetitive, time-consuming tasks the team had been doing by hand. The goal was to make the system do the coordination work, so the team could focus on the parts of the job that actually required human judgment. An AI agent was deployed on the website to handle inbound FAQs, qualify leads before human contact, and surface the right information without anyone having to intervene.

Key deliverables


Results

<2hEnquiry response time, down from 3+ days — handled automatically from the moment of form submission
100%Lead follow-up rate via automated sequences, up from an estimated 30% with manual handling
~12hTeam hours saved per week — freed from manual coordination, copy-pasting, and chasing maintenance updates

Skills applied

AI automation Agentic workflows Workflow automation AI agent deployment HubSpot CRM Zapier Operations audit Process design Lead nurture design Tool integration Property management operations