Welcome, everyone, to the Straker Limited half-year results presentation. I'm Grant Straker, the co-founder and CEO, and with me is David Ingram, our Chief Financial Officer. This presentation should take around 12 minutes, and we'll then open the floor for questions. If you have any questions during the presentation, please put them into the Q&A chat, and we will answer in the Q&A section at the end. We are here to discuss our H1 results, but the headline isn't just the numbers; it is the confirmation of our new model. The renewed NZD 28 million IBM contract for our core services gives a solid base for the next few years, and today I'm going to show you the blueprint for our AI-driven growth that our new partnership enables. Here is our narrative for today: we are executing a strategic planned pivot.
We have built an R&D engine we call the Small Language Model Factory. We have validated that technology with IBM, and most importantly, we are transitioning into a new model built for high-margin, scalable growth. Think of this as crossing a channel. On the left, you have the old world of legacy translation services. It's predictable but slow, low-margin, and relies on hiring more people to grow. On the right is the AI-native future. It is fast, high-margin, and scales with compute power, not headcount. It is also deployed through ecosystem partners, reducing our cost of sales. Our H1 results reflect the necessary investment to cross this bridge and secure our future with IBM and Watson X. Let's put the H1 performance into strategic context. We are sitting on NZD 8.7 million in cash with zero debt, which allows us to fund this pivot from a position of strength.
We are deliberately transitioning customers to our AI platform. While this impacts short-term revenue, it puts the building blocks in place for our new revenue foundation. Let's talk about the technology. There is a lot of hype around general large language models, MI, the jack of all trades, but enterprises find them expensive and slow. Our focus is the Specialized Small Language Model, SLM. These are experts, fast, cost-effective, and secure. Our proprietary TURI platform is essentially a Small Language Model Factory designed to build and deploy these high-value models at speed. The market is confirming our investment thesis. As the Wall Street Journal noted just last month, small models do the real work. Enterprises don't just want giant chatbots; they need specialized, efficient agents. We built TURI to deliver exactly what the market is now demanding. This brings us to the IBM partnership.
This isn't just a contract; it is a NZD 28 million validation of our technology. On top of the core business in the contract, we are co-developing specialized SLMs for IBM's Watson X platform. We have shifted from being a vendor to being a partner embedded in their ecosystem, which opens up massive low-cost distribution channels for us. Operationally, we are moving fast. We were at the IBM Innovation Center in Tokyo the first week of December, and we have integrated enterprise features into our Verify app to help with adoption. We are seeing pilot activity build, and we are currently exploring deploying Verify directly into the IBM Orchestrate catalog. We are also adding video translation features for SwiftBridge to meet key market requests, and while delayed, are seeing renewed interest. Why are we confident? Because we have achieved three layers of validation.
Technology, our SLM factory is proven at the highest level with IBM. This model, we have secure revenue foundation. Distribution, we are now embedded with a global AI hyperscaler. I want to address the two hardest questions head-on. First, margins. We are looking to build an AI business model that moves us from services margins to software margins, scaling with compute rather than humans. Second, timeline. The three-year IBM deal is the transition timeline. We are currently building the foundation with the full contract commencing January 1st, 2026. We see some great opportunities for fiscal year 2027, and in three years, we expect 100% of our revenue to be AI-derived revenue. Here is the roadmap. Phase one was building the people and tech. We are now in phase two, execution. Our job is to deliver flawlessly on the IBM co-development.
Looking ahead to phase three, we will scale this SLM factory model to other high-value verticals like legal and finance within the IBM catalog. To summarize, why Straker and why now? We are executing a strategic pivot backed by a NZD 28 million three-year blueprint. Our technology is validated by the market and by IBM. We are transitioning to a high-margin recurring revenue future, yet it's evident the market still sees us as a services company. Hopefully, the quantum and scale of the IBM deal can change this perception. Now over to David to go through the financial results.
Thank you, Grant. I'll now take you through our financial results for the half year. Despite a drop in revenue, we continue to deliver positive adjusted EBITDA. This performance reflects our focus on margin, operational efficiency, investment in research and development, and our evolving revenue model toward AI. Let's dive into the key highlights. Turning to the financial results for the half, revenue was NZD 19.3 million, down 15% on the prior period, which was consistent with the guidance we provided at the FY 2025 AGM. The decline was mainly driven by softer language services demand and continued attrition in our legacy enterprise TMS customer base. Despite this, our gross margin remained very strong at 66%, only slightly below last year. This reflects the benefits of workflow optimization and the early impact of our integrated AI tooling across production.
On costs, we continue to execute with discipline, underlying operating expenses reduced by 5% versus the prior period, reflecting a leaner production model, headcount efficiencies, and a shift toward lower-cost delivery regions. The increase in operating expenses as a percentage of revenue is simply a function of the lower revenue base rather than any cost expansion. We also continue to invest in the future. Capitalized software development increased 26% as we advanced our Verify platform and TURI Small Language Model programs. Adjusted EBITDA for the half was NZD 0.5 million, while lower than last year, we maintained profitability despite revenue headwinds. We also saw a substantial improvement in below-the-line items. Unlike the prior corresponding period, there were no impairment charges or amortization of acquired intangibles this half. Software development amortization increased modestly, in line with our expanding AI investment. Overall, our statutory net loss improved by 76% to NZD 1.3 million.
This was supported by strong cost control, a more stable currency environment, and a significant reduction in finance-related ethics losses. Taken together, these results demonstrate a resilient operating model, the impact of disciplined execution, and increasing leverage as we integrate AI deeper into our workflows. Revenue for the half year came in at NZD 19.3 million, down NZD 3.47 million versus the prior period and consistent with the guidance we provided. The largest movement was in language services, which was NZD 3.0 million lower, reflecting softer demand across several customers, especially government, where both budget constraints and pricing pressure affected volumes. Subscriptions were mixed. Traditional TMS subscriptions declined by around NZD 0.3 million due to ongoing attrition in traditional enterprise accounts, while Verify contributed a modest NZD 0.2 million uplift as early adoption continues to build.
Managed services revenue was NZD 0.4 million lower, driven by efficiency gains shared with our customer, an outcome that reflects the strength of our automation capabilities. Despite these headwinds, revenue performance remains in line with our full-year expectations, and we continue to shift the business toward higher-margin technology-led revenue streams. To illustrate the drivers of margin movement year on year, we analyzed EBITDA margin on a percentage of revenue basis, which cleanly isolates operating efficiency from the revenue decline. Starting with the half-year FY 2025, adjusted EBITDA margin of 7.3%, cost of sales increased slightly as a proportion of revenue, reducing margin by 0.7 percentage points. Selling and distribution expenses improved by 0.3 percentage points, reflecting operational efficiencies. Product design and development costs increased as a share of revenue, reducing margin by 2.5 percentage points.
General and administrative expenses increased by 3.8 percentage points, primarily due to fixed overheads being absorbed over a smaller revenue base. Other income, largely R&D tax credits, contributed positively by 2.2 percentage points. These movements resulted in a half-year FY 2026 adjusted EBITDA margin of 2.8%. After normalizing for restructuring and acquisition-related costs, the reported EBITDA margin for half-year FY 2026 was 2.5%. Overall, the walk highlights the increased weighting of product design and development and G&A costs on a lower revenue base, partially offset by efficiency gains in selling and distribution and the benefit of other income. Gross margins continue to perform strongly, increasing from 54% in fiscal year 2022 to 66% in the first half of fiscal year 2026. These gains reflect sustained efficiency improvements, greater automation, and a more technology-led mix.
Importantly, margins have remained stable at these elevated levels despite lower revenue, underscoring the resilience of the operating model. Turning to the balance sheet, our financial position remains solid. Cash ended the half at NZD 8.7 million, reflecting the timing of customer receipts and the refund of customer deposits, as well as continued investment in our AI development roadmap. Importantly, the business remains debt-free, giving us flexibility as we execute our strategy. Working capital reduced slightly by NZD 0.7 million, driven mainly by the unwind of prior-period customer deposits, while receivables and payables remained broadly stable. Overall, our working capital settings remain healthy and align with the current revenue base. Together, these movements reflect disciplined balance sheet management while continuing to invest in the technology assets that underpin future growth. That concludes the financial section. Back over to you, Grant.
Thanks, David. To outlook, I am pleased to reaffirm our guidance of revenue in the range of NZD 38 million-NZD 41 million and positive adjusted EBITDA. Now for any questions.
Hi everyone. Hopefully, you can hear me. We have a couple of questions just here. I have David just next to me. The first question is from Victor, which is, with the requirement for the Tokyo Stock Exchange to release announcements in English and Japanese from April 2025, should we have expected SwiftBridge to be up and running now and contributed materially to revenues?
It was mentioned that there are still some pilot activity at the moment. The Tokyo Stock Exchange mandated that from April 2025, there would be a requirement for companies to release in English and Japanese, but they do not actually enforce it until April 2026. What we have seen is customers waiting for that April 2026 when they actually will have some enforcement around that.
I'm actually flying up to Japan on the weekend, and we have a major week with IBM, actually, and SwiftBridge opportunities next week. We are very hopeful that over the next couple of months, we can start to show some really strong progress on SwiftBridge as that April 2026 deadline approaches. The next question, again from Victor, was we noticed a previous substantial holder in Clime Investment Management starting to sell down its holdings. Have they reached out and shared any feedback? We know that Clime is under a different investment manager at the moment in terms of the fund, and we have not seen any selling recently. They obviously did have a period where they were selling down, but we actually have not seen any for the last few weeks. At the moment, they are the two questions. Are there any other questions? Does not look like we have any other questions.
As always, feel free to touch base with either David or myself, and we're always happy to engage with shareholders on any specific questions you may have. That will conclude today's webinar, and we look forward to seeing everybody at the full year, if not before, if we have any other news to share along the way. Thank you.
Thank you.