Appen Ltd. · ASX:APX
Appen — Shareholder Meeting (May 2025)
34 slides from Appen's shareholder meeting presentation, May 16, 2025
Slide 1: 2025 Annual General Meeting – CEO address and presentation
ASX Release 16 May 2025 2025 Annual General Meeting – CEO address and presentation Appen Limited (ASX: APX) provides the attached CEO address and presentation to be delivered at today’s Annual General Meeting commencing at 10.00am AEST. Authorised for release by the CEO and Managing Director of Appe
Slide 2: CEO Address
CEO Address 2025 Annual General Meeting 16 May 2025 Opening Good morning everyone, and a warm welcome to Appen’s 2025 Annual General Meeting. It’s my privilege to address you as we reflect on a year of strong progress and outline our vision for the future. As a global leader in AI training data, App
Slide 3: CEO Address: Appen powers the world's best AI
Model architectures, such as transformers and diffusion models, are the important foundation for AI development. However, these are increasingly open source and accessible to anyone, meaning the competitive edge no longer lies solely in the model itself. Compute, powered by GPUs and TPUs, is largely
Slide 4: ADAP, CrowdGen and Mercury platforms
ADAP, our data annotation platform, provides a highly configurable interface for contributors to produce accurate annotations, supported by AI quality validators. We have a similar annotation platform that has been custom built for our China clients called MatrixGo. Our newly enhanced CrowdGen platf
Slide 5: Evolving data demands
Smaller pilots, generally under $100,000 and aligning with the important experimental nature of AI, accounted for 9% of our revenue. It’s worth noting that most of our large projects started as a small pilot. Together, these project types drove our 2024 performance, demonstrating our ability to oper
Slide 6: Strategic priorities and 2024 performance
• Second, we are evolving our data offerings to address the rapidly changing needs of our clients, supported by our flexible technology infrastructure. The AI market is changing rapidly, and we are building robust and flexible platforms with the ability to evolve with the market. • Third is technolo
Slide 7: Generative AI major driver, global product growth, opex reduced
Generative AI Major Driver Generative AI was a major driver of our growth, contributing 28% of our H2 2024 revenue. While non-LLM segments grew steadily at 2% half-on-half, the growth in generative AI projects reflects our ability to pivot to high-demand areas of the market. Global Product Growth Gl
Slide 8: Strategic priorities for 2025
• Second, we’re driving operational efficiency and automation by utilising AI across our operations. This includes further incorporating generative AI into our platforms for higher quality data delivery. • Third, we’re accelerating technology innovation by utilizing AI across our software developmen
Slide 9: Longer-term targets
Longer-term targets In relation to Appen’s longer-term targets we continue to be backed by solid market fundamentals and tailwinds. The opportunity for Appen is driven by large-scale investment in the development and utilization of generative AI, and reinforced by robust estimates for market growth
Slide 10: CEO Address
CEO Address Ryan Kolln © 2024 proprietary & confidential
Slide 11: Agenda
Agenda 1. Market Overview (incl. role of Appen) 2. 2024 Performance Recap 3. 2025 Guidance and Longer-term targets © 2024 proprietary & confidential 2
Slide 12: Market Overview (incl. role of Appen)
Market Overview (incl. role of Appen) © 2024 proprietary & confidential
Slide 13: AI remains driven by a combination of model, compute & data
AI remains driven by a combination of model, compute & data Model Compute Data Deep learning (Convolutional Hardware: GPUs / TPUs Public data (e.g, internet Neural Networks) scraping) Generative AI (Transformers, Proprietary data including diffusion network, large bespoke datasets language models) ©
Slide 14: Appen enables the world's best AI through high quality human data
Appen enables the world's best AI through high quality human data © 2024 proprietary & confidential
Slide 15: Appen creates bespoke AI training data for leading model builders
Appen creates bespoke AI training data for leading model builders Client requirements Appen AI data consultants Description of task Workforce requirement Workforce and project management Bespoke AI Data annotation platform: ADAP training data platform: Mercury Data quality rubriks Volume of data req
Slide 16: Scalable and flexible technology underpins Appen's operations
Scalable and flexible technology underpins Appen’s operations Proprietary technology platforms that enable high quality data at scale Appen AI data consultants Mercury ADAP + MatrixGo1 CrowdGen Crowd and project management platform Data annotation platform Crowd onboarding and interface Team of expe
Slide 17: Examples of how Appen supports its customers
Examples of how Appen supports its customers Evaluation of LLM Ad evaluations Image annotation responses Foundation Social Misinformation Domain specific AI first Embodied AI model media categorization model reasoning companies agents (robotics) builders AR/VR & wearable Agentic AI training Multimod
Slide 18: Data requirements range from continuous feed to small and experimentative
Data requirements range from continuous feed, large scale and finite, to small and experimentative % of Appen Project characteristics Examples of customer requirements revenue in 2024 Continuous data feed • Solving for real-world changes >6 months duration • 51% 51% projects Ongoing improvements to
Slide 19: Generative AI data demands are evolving rapidly
Generative AI data demands are evolving rapidly, with a trend to more complex and shorter project durations Workforce Projects are more Task complexity is requirements are experimentative and increasing more specific shorter in duration • Reasoning and red-teaming tasks can • Level of expertise in t
Slide 20: Case study of recent LLM evaluation project
Case study of recent LLM evaluation project High volumes with short delivery timeframes Project overview: large scale evals 550 contributors matched Rapid ramp within hours with rapid scaling across 10+ domains of project launch Units of data delivered in 4 days Client request was to evaluate 100,00
Slide 21: Appen sees three critical areas for long-term success
Appen sees three critical areas for long-term success Expert-led approach Evolved data Technology enabled to sales and offerings efficiencies marketing • Appen customers are at the forefront • The data needs of customers is • Creating high quality data increasingly of AI model development evolving r
Slide 22: 2024 Performance Recap
2024 Performance Recap © 2024 proprietary & confidential
Slide 23: 2024 was a pivotal year for Appen – balancing cost and growth
2024 was a pivotal year for Appen – balancing cost and growth Broke into LLM market, with 28% 16% revenue growth for the full Grew China revenue by 71% of revenue in H2 FY24 from LLM year, excluding Google projects Returned to profitability, with Appen technology playing a Reduced full-year OPEX by
Slide 24: Revenue growth returned
Revenue growth returned Group revenue Commentary • Excluding impact of Google, Appen experienced a 80.0 return to revenue growth largely due to the rise of generative AI 70.0 • Excluding Google, revenue for Q4 FY24 grew 37% 19.1 60.0 23.3 on pcp (from $48.6 million in Q4 FY23) 18.3 13.4 • Growth con
Slide 25: China success continued
China success continued China division revenue Commentary 20.0 • 71% growth in FY24 compared to FY23 • China customers include leading LLM model 18.0 builders, along with leading technology and auto customers 16.0 • Projects are mostly for in-facility annotation, 14.0 resulting in a more predictable
Slide 26: Generative AI major driver of revenue growth
Generative AI major driver of revenue growth 28% of group revenue from generative AI-related projects1 Non-generative AI projects growing ~2% HoH in H2 FY24 30% 140 28% 120 33.6 100 20% 14.7 5.3 80 15% 60 10% 83.5 85.3 87.2 40 6% 20 0% H2 FY23 H1 FY24 H2 FY24 0 H2 FY23 H1 FY24 H2 FY24 Other revenue
Slide 27: Global product growth reflects reliance on Appen's technology
Global product growth reflects reliance on Appen’s technology Global Product revenue Commentary 16.0 • Global Product revenue represents projects that are delivered on Appen’s data platform 14.9 14.0 • For Appen’s largest US technology customers, most work traditionally delivered on their internal 1
Slide 28: Opex reduced by 37% H1 FY23 to H2 FY24
Opex reduced by 37% H1 FY23 to H2 FY24 Opex1,2 Commentary 80.0 • Employee expenses reduced by 41% and all other expenses by 28% compared to H1 FY23 • Significantly reduced product and engineering spend achieved by establishing a hub in Hyderabad 60.0 20.0 • Consolidation of business units and ration
Slide 29: EBITDA improving as costs remain in control
EBITDA improving as costs remain in control Underlying EBITDA before FX1 Commentary • Underlying EBITDA1 (before FX) for Q4 FY24 was a 8.0 profit of $4.7 million, continuing the return to profitability 6.0 • Acted swiftly to control costs following the Google 4.0 4.7 announcement and completed $13.5
Slide 30: 2025 Guidance and Longer-term targets
2025 Guidance and Longer-term targets © 2024 proprietary & confidential
Slide 31: Near-term strategy
Near-term strategy Phase 1 of Appen’s 12-month strategy focusses on delivering improved data quality and speed (unit economics), further bolstering Appen’s core business and greater revenue diversification Enhanced focus on LLM growth Operational efficiency & automation Accelerate technology innovat
Slide 32: FY25 Guidance
FY25 Guidance The Company is optimistic about its FY25 revenue opportunity, supported by: • More-predictable project work from large US tech customers, typically H2-skewed; • Sustained growth in China; and • Work not yet included in pipeline, including less predictable generative AI related work, ty
Slide 33: Longer-term targets
Longer-term targets In relation to Appen’s longer-term targets (FY27+), the Company continues to be backed by solid market fundamentals and tailwinds. The opportunity is driven by large-scale investment in the development and utilization of generative AI, and reinforced by robust estimates for marke
Slide 34: Thank you
Thank you © 2024 proprietary & confidential 25