IIIV AI Software Development Boosts Velocity Over 25%

On August 7, 2026, i3 Verticals, Inc. (IIIV) said on its third-quarter earnings call that its internal AI software development tools increased development velocity by more than 25%. The company was releasing software more frequently while keeping engineering headcount flat, showing higher engineering capacity, but it did not separately disclose the contribution to operating profit.[1]

i3 Verticals (IIIV) provides vertical software to the US public sector. Its internal product, engineering, quality assurance, cloud operations and security teams use AI to plan, write, review, test and modify code and to deliver product features and customer system integrations. The tools are part of the core development and delivery process. They improve engineering output rather than serving as a separately priced product sold to customers.

How IIIV's AI development workflow evolved

The internal software development and delivery application has evolved continuously within the same workflow. On May 9, 2025, management first cited AI alongside offshore development as a factor offsetting software investment costs, allowing the company to increase output without substantially increasing costs.[2] Three months later, the company named GitHub, Copilot, Builder IO, Cursor, CodeRabbit and Zephyr across coding, review and testing, and said development-process efficiency improved by 30% to 50%, depending on the team and the code.[3]

By May 2026, the workflow had advanced from suggesting code snippets to using AI agent tools that could plan, write, test and modify code with minimal human intervention.[4] In the latest disclosure on August 7, 2026, the agentic workflow had reached a stage where the company measured it through internal sprints and project delivery results.

What the more than 25% velocity gain shows

The most direct current result comes from internal sprint metrics: development velocity improved by more than 25%, software releases became more frequent and engineering headcount remained flat.[1] This means the same workforce can handle more features and releases. The company did not disclose the baseline period, sample size, measurement window or absolute engineering headcount, however, so the figure cannot be used to calculate avoided hiring or expense savings.

A customer court case management system integration provides a specific delivery example. The company said the project took approximately one-third the time that similar projects historically required and that AI-assisted development produced a reusable, configurable integration framework.[1] The result covers only one project, while the historical comparison baseline and absolute project duration were not disclosed. It therefore cannot be extrapolated into an average companywide development cycle. The company also stressed that human oversight with domain expertise and established quality controls remain part of the process.[1]

The financial impact remains unquantified

If development velocity and reuse continue to improve while engineering headcount remains flat, labor input per feature, release and customer integration could decline and eventually support operating profit. However, adjusted EBITDA margin for the quarter was 25.0%, compared with 24.5% a year earlier. Management attributed efficiencies and savings jointly to process improvements and AI adoption. That measure is not operating profit, and it combines this workflow with other AI applications as well as changes in personnel, rent costs and other factors, so the contribution from the internal development workflow cannot be isolated.

The verified change is that i3 Verticals has advanced AI from a development assistance tool into an operating capability that performs multi-step tasks and is measured through sprint and project results. To establish the financial return, the company would still need to disclose engineering cost per release or customer integration on a comparable basis and separate the effects of AI, capitalized software investment, personnel costs and other process improvements.

Application assessment

• AI-Assisted Software Development and Delivery | Business position: Core operations | Deployment stage: Limited production | Scope: Companywide | Value type: Cost reduction

Sources

[1] Drillr · i3 Verticals, Inc. (IIIV) · 2026-08-07 · Earnings call

Original: Internal sprint metrics show that more than 25% improvement in development velocity and we are releasing software more frequently while maintaining a flat engineering headcount.

[2] Drillr · i3 Verticals, Inc. (IIIV) · 2025-05-09 · Earnings call

Original: So even though we're kind of increasing investment in the rate of CapEx and what we're developing on the software side, there is things kind of cutting the other direction on that too between AI offshoring, we're kind of increasing our output without having to increase our cost substantially.

[3] Drillr · i3 Verticals, Inc. (IIIV) · 2025-08-08 · Earnings call

Original: While it depends on the team and the code, our engineering team have found 30% to 50% more efficiency in their dev process when incorporating AI.

[4] Drillr · i3 Verticals, Inc. (IIIV) · 2026-05-08 · Earnings call

Original: We began with AI assistance across the enterprise, which in layman's terms simply suggests snippets of code to enhance overall code development, debugging, testing, among other uses, and now have moved to AI agent tools that plan, write, test, and modify code with minimal human intervention, and it increased our product development capabilities, allowing us to pursue opportunities that would have previously required difficult trade-offs and prioritization, such as new feature development and product releases.