Company Spotlight: Trebellar Builds AI for Corporate Real Estate

Corporate real estate runs on physical commitments and moving assumptions. A company may sign a lease for years while headcount, office policy, commute behavior, costs, and employee expectations change by the quarter.

Trebellar builds AI software for that mismatch. The San Francisco company connects lease, headcount, cost, space, utilization, market, and employee-experience data in a corporate-real-estate-specific ontology, then gives AI agents a governed record from which to plan, analyze, reconcile, and report.

Co-founders Diego Ferreiro Val, CEO, and David Garcia Quintas, CTO, started Trebellar in 2021 after a home-automation side project exposed a larger enterprise problem. Corporate real estate teams had plenty of software and data, but portfolio decisions still depended on assembling the relationships between systems.

Trebellar matters now because enterprise AI is moving from broad assistants toward systems that understand one operating domain deeply enough to earn a place in consequential work. The company's challenge is to make the portfolio legible without pretending that software should own the human judgment behind a lease, consolidation, relocation, or workplace policy.

About Trebellar

Trebellar is an AI-native corporate real estate platform headquartered in San Francisco. Its stated mission is to expand the use of AI among real estate leaders, with software designed around how corporate portfolios are planned, financed, occupied, and changed.

The company's origin began during the pandemic, when Ferreiro Val and Garcia Quintas were experimenting with a centralized home-automation application. A conversation with Salesforce's global workplace technology leadership redirected the idea toward enterprise buildings, where data was distributed across point systems, dashboards, documents, and spreadsheets.

That shift gave Trebellar a specific design premise: a real estate decision lives in the relationships between records. A floor sits inside a building, a team occupies the floor, a lease governs the space, cost attaches to the lease, and utilization changes the economic meaning of the commitment. The product has to preserve those relationships before an AI agent can offer a useful answer.

The Ontology Under the Agents

Trebellar's corporate real estate ontology connects leases, buildings, floors, teams, costs, headcount, occupancy, utilization, market information, HRIS data, surveys, and operational systems. The company supports CSV uploads, secure file sync, APIs, warehouse connections, PDFs, and other sources, then normalizes the inputs into one governed model.

The model is the foundation for Trebellar's functional and domain agents. Functional agents can coordinate plans, alerts, reports, and summaries. Domain agents bring context for lease administration, transactions, portfolio strategy, and analysis. A functional agent can request specialized analysis, assemble it into a report, and expose the source records and assumptions used in the result.

Trebellar's product materials make a deliberate boundary around automation. Agents gather, reconcile, and draft. People decide the shape of the portfolio, negotiate with landlords and business units, and choose which trade-offs are acceptable when data points in different directions. That separation is central to the commercial proposition because a confident answer has little value if the executive cannot inspect how it was produced.

Product Surface and Enterprise Readiness

The Trebellar platform includes portfolio management, planning and strategy, lease optimization, location research, workplace experience, dashboards, Explorer, Portfolio Planner, reports, and connections for outside AI tools through MCP. Trebellar's Explorer adds market, talent, transit, commute, amenity, cost, utilization, and sentiment context to location decisions.

The company says a CSV-based start can take as little as 2 to 3 weeks, while direct warehouse or full API integrations generally take 2 to 6 weeks. Those timelines are company guidance rather than a service guarantee, but they show how Trebellar frames adoption: begin with a governed subset of the portfolio, then deepen the integrations as the operating record becomes more valuable.

Enterprise readiness also depends on governance. Trebellar's security page lists SOC 2 Type 2 and GDPR, TLS 1.2 for data in transit, AES-256 at rest, SSO, MFA, user permissions, regular backups, vulnerability scanning, static code analysis, and masking when integrated data contains personal information. These are published company controls; the underlying audit report was not independently reviewed for this profile.

Leadership, Customers, and Capital

Ferreiro Val previously led engineering work at Salesforce. Garcia Quintas built AI and infrastructure systems at Google and Waymo. Trebellar's current leadership page also lists Adam Sewall as CRO/COO, joining technical product leadership with the enterprise sales and operating work required to move the platform through security review, integration, and procurement.

Trebellar names Meta, Uber, Merck, and Cohesity as customers in its September 2026 financing announcement. Lead investor Blossom Capital also lists Workday and Okta. Those names indicate that Trebellar is being tested inside complex enterprises, although the company has not disclosed revenue, retention, a verified customer count, or independently audited savings tied to those deployments.

The company raised an $18M Series A led by Blossom Capital on September 24, 2026. Haystack, Alt Capital, 1Flourish, and Bynd Venture Capital participated. Trebellar said the capital will expand engineering and go-to-market, deepen its AI capabilities, and bring the platform to more enterprises.

Culture and Hiring as a Market Signal

Trebellar's careers page describes a culture built around integrity, transparency, curiosity, accountability, sustainability, and humility. The company lists remote, office, and hybrid options, along with medical and wellness benefits, flexible time off, and work on product and built-environment problems.

As of September 24, 2026, Trebellar's public openings include Senior UX Designer, Senior Front End Engineer, and Senior AI/ML Engineer roles. That mix is a useful signal. Trebellar is investing in the interface where operators interrogate portfolio data, the application layer that delivers the work, and the models that power forecasting, reconciliation, and agent behavior.

Hiring cannot establish traction by itself. It does show which capabilities management believes must deepen next. For Trebellar, the open roles sit across the exact handoff the company has to get right: sophisticated data and AI underneath, an enterprise-grade product in the middle, and an interface that a real estate leader can use when the answer carries a long financial tail.

What Trebellar Signals for Enterprise AI

Trebellar represents a broader shift from generic enterprise assistants toward systems built around governed domain models. Corporate real estate is a useful proving ground because the records are fragmented, the relationships are technical, and the decisions affect both cost and people over long periods.

The company's durable advantage will depend on how well its ontology survives messy integrations, changing definitions, disputed records, and the pressure of real lease events. Each successful workflow can make the shared model more useful across planning, operations, and reporting. Each bad source or opaque assumption can weaken trust across all of them.

That is the work waiting behind Trebellar's polished product surface and new capital. The company is building for the point where an enterprise stops asking whether AI can produce an answer and starts asking whether the organization can defend that answer after the lease is signed and the portfolio has moved again.