We've raised a $15M Series A led by Accel, with continued participation from Y Combinator, bringing our total funding to over $20M. With 500+ customers, 2+ years of profitability, and nearly 3x year-over-year recurring revenue growth, we're scaling AI for construction financials, deeply integrated with the systems contractors use today.

Construction runs on decades-old software

We serve the contractors who build the physical world around us: the data centers, transit hubs, and hospitals we depend on every day. It’s easy to see the cranes and scaffolding soaring high on jobsites, but humming behind every job is a hidden world of unglamorous, essential work. This is where our customers make or break their margins, and where we plug in.

Every month, construction accounting teams do an enormous amount of work to keep jobs compliant and profitable. They process thousands of invoices, track change orders, manage retainage, reconcile job costs, forecast cashflows, bill customers, produce reports, and more. This work is disjointed and manual, done through emails, paper, spreadsheets, and software built decades ago.

To anyone in construction, this story will sound familiar: each month, a supplier sends invoices in a two-hundred-page PDF. Someone in accounting parses it page by page. They figure out which invoices belong to which project, and route copies to project managers for approval. Costs are logged in the ERP, but they don’t push over to the project management system. Job-cost reports are four weeks outdated, and the first week of every month is spent reconciling data to close the books. Until then, no one trusts the numbers they're seeing, and project financials are based on intuition.

This isn't because contractors resist new tech. Their work requires compliance with industry rules and specialized processes that newer, more-generic software doesn’t support. Migrating systems means hundreds of thousands of dollars in fees and quarters of change management. A single hiccup can mean missing payroll, failing an audit, or not paying vendors. Any one of these threatens the decision-maker’s job and the company’s reputation. It’s no surprise that contractors opt to stay on existing systems, which leaves thousands of companies running on software older than their employees.

Building the industry's infrastructure

We spent the past four years building the plumbing to connect the most vital software in construction: systems covering accounting, project management, scheduling, and more. Most of these have no public APIs and no documentation, and many are still on-prem.

Connecting them required going into their guts, understanding hundreds of workflows per system, abstracting their data models, reconciling their differences, and learning each software’s quirks, one at a time.

We packaged our work in a unified API for construction software vendors that made it easy to move data between systems. We scaled to over 12K contractors using our API. GCs and Subs then asked to use our API to keep their systems in sync, and to provide real-time financial data to field teams on their most critical projects. Today, contractors across North America have used Agave on over 80K projects representing over $100B in construction volume.

We're proud of the jobs we’ve powered so far: the Tesla Gigafactory, data centers for AWS and Meta, the new light rail in Austin, Lumen Field in Seattle (where the Seahawks play), dozens of hospitals, universities, and even the WWII Memorial in DC.

Becoming the trusted partner for back-office AI

As we worked closely with hundreds of customers, they asked us to solve a growing range of problems. We watched customers hack together partial solutions using consumer AI tools, but struggle with the same set of gaps: data access across systems, construction specificity, permissions, customizations, security, and more.

We shipped our first AI product in late 2025 to fill these gaps, powered by our own unified API. Returning to the example above, our AP Invoice agent fetches that two-hundred-page supplier PDF, splits it into separate invoices, compares the invoice to relevant contracts and flags variances, codes each line to the right cost code, routes it for approval, and writes validated data back to the contractor’s ERP. Within months, we launched three more AI products. Since then, recurring revenue from our AI products has grown more than 10x, back-office teams are getting hours of their month back, and operations teams finally have real-time financials they trust.

As two examples:

• RW Dake Construction more than doubled in size while keeping its accounting team lean and making them more productive. Agave keeps their Procore and Foundation data in sync automatically, and their CFO uses our AI to do in minutes what used to take days: splitting a master job report into tailored financials for each of 47 project managers, or scanning months of credit-card activity to catch expenses coded to the wrong job. Read their story here.
• Brinkman Construction cut monthly forecasting from days of spreadsheet work to hours, at a quarter of the labor, and now prices change orders before work starts instead of absorbing owner-driven costs after the fact. They're rolling out our AP Invoice agent to replace hand-coding hundreds of invoices every month. Read their story here.

This is why our AI is different. Most AI products are limited in their ability to access enterprise data, understand the business and domain context, and act on that information to get stuff done. Our proprietary data access and read/write capabilities across all industry platforms make our tools uniqueful useful, especially when operating on construction financials where data spans multiple systems, accuracy has to be near-perfect, and results need to be easy to audit.

What's next

Construction is losing the people who hold its financial know-how. The National Center for Construction Education and Research (NCCER) projects 41% of construction’s workforce will retire by 2031, and only about one in ten workers today is under 25. This shift creates an urgent challenge for one of the largest industries in the world. It’s a major opportunity for those able to help.

We see AI as the highest-leverage way contractors close that gap, by capturing and applying organizational knowledge so it doesn't walk out the door. Accelerating this opportunity is why we raised. We've been profitable for over two years and didn't need the money to keep going. We raised because demand for AI in construction back offices is outpacing what our team could build, and moving fast matters more than funding it with our own profits. We’re investing the new capital into two areas: 3x-ing our team size this year, and doubling the breadth of our product suite.

Our goal is simple: build the AI financials platform that every contractor in North America runs alongside their existing systems. We're shipping 4+ new products in the next 6 months, which means the people who join now own entire products end to end. If you want to work on hard problems that shape the physical world, from our office in downtown SF, we'd love to meet you: see our open jobs here.