Automating bill processing, lease intelligence, and tenant invoicing for a South Florida real estate boutique

The client

A real estate boutique in South Florida with over 15 years of experience, 350+ investors served, and 900+ commercial and residential units under management. The property management arm handles day-to-day operations across the portfolio: tenant relations, rent collection, maintenance coordination, vendor payments, lease management, and deal evaluation. A team of 12 running operations across multiple properties with overlapping vendors, varying lease terms, and a steady flow of new investment opportunities.

The systems

The firm runs on Propertyware® for property management, an accounting platform for general ledger and expense tracking, Gmail for incoming vendor bills and deal documents, and bank feeds for payment reconciliation. Before this project, none of these systems talked to each other through AI. Data moved between them manually, one entry at a time.

The problem

Every bill that came in had to be manually reviewed, classified, and assigned to the correct property and general ledger account. The accounting team read each invoice, identified the vendor, figured out which property it belonged to, and entered it into the system. With 1,000 vendor bills arriving per month across the portfolio, misclassifications happened regularly. A maintenance invoice for one building would get booked to another. A utility bill would land in the wrong GL account. Each error cascaded into reconciliation issues downstream.

Offer memorandums for real estate transactions arrived as unstructured documents. Each one had to be read, key terms extracted manually, and the relevant data entered into the tracking system. A single offer memo could take 30 to 45 minutes to process. During active deal periods with multiple offers coming in simultaneously, the team fell behind.

Tenant invoicing was equally manual. Lease terms, rent amounts, CAM charges, and billing schedules lived across Propertyware and spreadsheets, but generating and sending each invoice still required someone to pull the right numbers, create the document, and send it. For a portfolio with different billing cycles and variable charges, this consumed 15 to 20 hours every month.

What we built

We deployed 12 AI skills across 4 connected systems, organized into three core workflows. Each skill handles a discrete piece of the operation and feeds its output into the next.

Intelligent bill processing (5 skills)

• ✓Bill intake reads incoming invoices from email and uploads, extracts the document content regardless of format: PDF, image, or scanned copy.
• ✓Expense classification categorizes each bill by type: maintenance, utilities, insurance, taxes, management fees, capital improvements, and more.
• ✓Seller identification matches the vendor against the existing vendor registry or flags new vendors for review.
• ✓Property matching determines which property in the portfolio the bill belongs to, using vendor history, address references, and account context.
• ✓GL account assignment maps the classified expense to the correct general ledger account based on the expense type and property.

The result: a pre-filled expense entry with seller, property, classification, and account already assigned. The accounting team reviews and approves instead of building from scratch. Estimated time savings: 10 to 15 minutes per bill, multiplied across 200+ monthly bills.

Lease intelligence (4 skills)

• ✓Lease parsing reads active lease agreements and extracts key terms: rent amounts, escalation clauses, renewal dates, maintenance responsibilities, and special conditions.
• ✓Renewal tracking monitors upcoming lease expirations and flags properties that need attention before deadlines.
• ✓Maintenance responsibility lookup allows property managers to ask who is responsible for a specific repair based on the lease terms for that unit.
• ✓Lease Q&A enables property managers to ask natural language questions about any lease: when does it renew, what are the CAM charges, who handles HVAC maintenance.

Property managers went from digging through lease files to asking a question and getting an answer. Renewal dates are tracked automatically. Maintenance disputes are resolved by reading the lease instead of guessing.

Automated tenant invoicing (3 skills)

• ✓Lease term tracking monitors active leases in Propertyware, including rent amounts, CAM charges, escalation clauses, and billing schedules.
• ✓Invoice generation creates accurate invoices for each tenant based on their specific lease terms, billing period, and any outstanding balances or adjustments.
• ✓Delivery scheduling queues invoices for the correct send date based on each property's billing cycle, ensuring tenants receive invoices on time regardless of varying schedules.

Monthly invoicing went from a 15 to 20 hour manual process to a review-and-confirm workflow. No missed billing cycles. No incorrect amounts. No scrambling at month-end.

The results

The accounting and operations team went from spending the majority of their time on data entry to reviewing AI-prepared outputs. Bills are classified correctly, lease questions are answered instantly, and tenants receive accurate invoices on schedule without anyone manually creating them. For a 12-person firm managing a growing portfolio, the recovered 240 hours per month transformed how the team spends their time.