Inmobil IA: AI & Automation Master Plan 2026–2028

Between January and December 2026 (first phase of a 2026-2028 roadmap), automate with Artificial Intelligence at least 12 recurring, low-value operating processes across Finance, Treasury, Accounts Payable, Sales/Marketing and Learning & Development, freeing the equivalent of 450+ person-hours per month currently spent on manual work - without growing the 7-person IT team or investing in physical infrastructure (0% CAPEX, 100% OPEX) - so operating capacity can scale with business growth, including expansion into two additional countries, without a proportional increase in organizational headcount.

As of September 2026, the first phase is already producing measurable results. Trust/loan disbursement packages for the bank now take under 5 minutes to assemble, down from roughly 30 minutes of fully manual work. The real-time inventory agent has recovered an estimated 840 hours per year of manual reporting capacity across all active developments. The commercial AI agent now delivers a first response to a new sales lead in under 5 minutes, compared to a process that previously depended entirely on staff availability. The new semantic data layer already supports roughly 90% of executive reports without an intermediate spreadsheet, up from about 20% before the project. The AI learning academy has been operating since June 2026, and cybersecurity awareness campaigns are sustaining a phishing-simulation pass rate above 95%. Purchase-order processing, launched in August 2026, is cutting per-document handling time from about 2 minutes to under 10 seconds. Beyond individual metrics.

Microsoft 365 (Teams, Outlook, Word, Excel, PowerPoint, SharePoint); Microsoft Power Platform (Power Apps, Power Automate, Power Virtual Agents, Power BI, Power Pages); Microsoft Copilot; Microsoft Azure; HubSpot CRM; Claude (Anthropic) AI agents; the company's ERP / real-estate back-office system; a contract-intelligence tool for legal, tax and accounting review; a Learning Management System for the internal AI academy; and a cybersecurity stack (endpoint protection, anti-phishing, backup).

The main complexity was organizational, not technical: delivering 22 simultaneous initiatives across 18 internal areas and multiple external partners with a corporate IT team of only seven people, while keeping a zero-CAPEX, 100%-cloud constraint that ruled out on-premises shortcuts. Legacy systems had to be integrated without a common data model, requiring a semantic data layer to be built before most Data Core and reporting initiatives could proceed - a single dependency that conditioned the rest of the program's timeline. Rolling out AI tools to roughly 200 employees with uneven digital literacy required a dedicated, mandatory training and change-management track, run in parallel with technical delivery. Finally, the plan had to be designed for a moving target: an active multi-country expansion that continuously changes transaction volume, regulatory requirements and the scale each solution must support.

Context and problem. The organization is a real estate developer with 30+ active residential and commercial developments, expanding from one country into two additional markets. Its operating structure had not evolved at the same pace as its business volume: a seven-person IT team supported technology needs across 18 internal areas, creating growing dependence on individual availability for recurring, low-value tasks. The clearest symptom sat in financial close: core systems (ERP, CRM, banking, internal controls) were not integrated, so reconciliation relied on manually built spreadsheets consuming roughly 450 person-hours a month. Assembling a trust/loan disbursement package for the bank took about 30 minutes, fully manual; purchase orders were captured one by one in Accounts Payable; and response to new sales prospects depended entirely on staff availability, producing one to two touchpoints per lead across 600+ monthly leads. No unified data layer existed and there was no formal AI training program. Scaling by adding headcount was neither viable nor financially responsible.

Solution implemented. We designed an AI & Automation Plan built on five pillars - AI for All, Autonomous Operations, Zero Trust, Data Core and Commercial Engine - underpinned by an internal AI-agent platform, an Automation & AI Office, and a mandatory change-management workstream. The first phase is already in production: a conversational, real-time inventory agent covering all developments; automated generation of trust/loan instruction packages; an AI agent that responds to and qualifies sales prospects within minutes; a real-time semantic data layer and executive BI; and a company-wide AI learning academy with role-based paths. The model runs on cloud infrastructure only - zero capital expenditure, 100% operating expense.

Difference and resolution. Unlike the previous model - manual execution by specific individuals, knowledge held in personal memory - the solution moves recurring operations to AI agents and automated workflows that are consistent, auditable and scalable, so capacity no longer depends on hiring to grow. It attacks the root cause: dependence on human availability for low-value work. Automating the trust-package process (30 to under 5 minutes) and prospect response frees hundreds of hours monthly for analysis, without enlarging the 7-person IT team that, supported by business "champions," now sustains 22 initiatives across 18 areas.

Innovation and beneficiaries. This is a transformational innovation: it redesigns the operating model - from manual, people-dependent operations to an intelligent, automated, scalable one - with incremental gains within each pillar. Direct beneficiaries: Finance, Treasury and Accounts Payable; Sales and Marketing, with faster attention to ~615 monthly leads; ~200 employees, who gain AI training and tools; external customers, with faster responses; and leadership, with reliable, real-time information for decisions

Mexico (current operations); expansion underway to Spain and Belize.