Arvind Ramnani

Arvind Ramnani asked how Globant plans to prevent cannibalization of its core headcount-based revenue as AI Pods, currently a small but growing part of revenue, scale up. He also inquired about internal modeling for the revenue crossover point between token-based and headcount-based revenue, and whether the margins on tokens are higher or lower than the company average.

Answer

CEO Martín Migoya stated that Globant actively encourages the transformation to AI Pods, viewing it as essential for future growth and scalability, leveraging years of experience into 'recipes' for enterprise-ready assets. CFO Juan Urthiague added that the model is accelerating, with some customers migrating $10M-$20M engagements, expecting AI Pods to be more relevant by year-end and curves to converge next year. Martín Migoya confirmed AI Pod margins are 45%-60%, higher than the blended 38%, and are expected to increase with efficiency, creating a virtuous cycle of productivity and value.

Arvind Ramnani asked about the strategy to prevent cannibalization of Globant's core headcount-based revenue as the AI Pods business, currently less than 1% of total revenue, scales. He also inquired about internal modeling for the revenue crossover point between token-based and headcount-based revenue, and whether AI Pods' gross margins are higher or lower than the company average.

Answer

CEO Martín Migoya stated that Globant aims for this transformation, viewing it as a rational migration that positions the company for AI-driven growth through scalable, repeatable, and enterprise-ready AI Pods. CFO Juan Urthiague noted the rapid growth of AI Pods from zero to $20 million ARR in two quarters, with increasing customer adoption for larger deals, expecting it to become more relevant by year-end and curves to get closer next year. Migoya confirmed AI Pods' gross margins are between 45%-60%, higher than the blended 38%, and are expected to increase with efficiency and improved supervision technology, creating a virtuous cycle of productivity and margin expansion.

Arvind Ramnani asked about the potential drivers for upside or downside to the revised guidance and questioned the implied sequential growth improvement from Q1 to Q2 given the volatile macro environment.

Answer

CFO Juan Urthiague reiterated that the new guidance is conservative and assumes current uncertainty persists. Upside could come from a clearing macro environment that accelerates pipeline conversion, while downside would stem from further deterioration. He explained that the Q2 guidance reflects a stabilization that the company is currently observing, and the number was set to be achievable in the current volatile scenario.

Arvind Ramnani highlighted a perceived disconnect between management's highly positive commentary and a guided growth rate that isn't significantly accelerating from the previous year.

Answer

CFO Juan Urthiague addressed this by stating that the guided organic constant currency growth is consistent with 2024's strong performance. He explained that the excitement stems from the transformative potential of AI, an expanding U.S. pipeline, and major wins in new markets, but acknowledged that deal conversion must accelerate for this to translate into higher revenue. CEO Martín Migoya added that achieving this level of growth at their scale is a top-tier industry performance and demonstrates their readiness for the future.

Arvind Ramnani asked if advancements from frontier AI models like Anthropic and OpenAI pose a competitive threat, or if CCC's industry-specific dynamics and unique data provide a strong defense. He also inquired if CCC works with these frontier models and if this pressures costs.

Answer

CEO Githesh Ramamurthy explained that customers already work with these companies for horizontal applications, while partnering with CCC for deep vertical expertise in claims, viewing it as an 'and' rather than 'either/or' scenario. He confirmed that CCC works with seven different companies, including frontier model providers, and has been doing so for several years. He highlighted CCC's decade of expertise in building and deploying AI, including sophisticated approaches to hardware, training, and inference models, which helps manage costs effectively.

Arvind Ramnani asked if advancements from frontier AI models like Anthropic and OpenAI pose a competitive threat, given that customers might partner with them, or if CCC's industry-specific data and dynamics provide a unique competitive moat. He also inquired about CCC's collaboration with these frontier models and any resulting cost pressures.

Answer

Githesh Ramamurthy, CCC's Chairman and CEO, clarified that customers already work with these companies for horizontal applications, and CCC itself uses these tools internally. He emphasized that CCC's deep vertical expertise, proprietary hyperlocal data, and embedded workflows for auto and casualty claims differentiate its offerings. Mr. Ramamurthy confirmed that CCC works with seven different companies, including frontier model providers, and that its decade of expertise in AI deployment, including sophisticated approaches to hardware, training, and inference models, helps manage costs effectively.

Arvind Ramnani asked how broader advancements in AI enable Upstart to leapfrog its internal AI programs and if any previously special developments are now more broadly available, potentially exposing Upstart. He also inquired about Upstart's key AI partners.

Answer

Paul Gu, Co-founder and CTO, expressed gratitude for AI progress, seeing it as a huge benefit, not a threat. He explained that general AI excels at human-like tasks, whereas loan underwriting is a complex math problem requiring specific data and proprietary algorithms Upstart has built over 12 years. Upstart benefits from basic infrastructure advances (chips, compute) and uses general AI (LLMs) for automating messy, human-centric operational processes like HELOC lien perfection, boosting automation rates. He clarified that while they use models from various foundation model providers for operational tasks, their core lending models are proprietary and not built directly on others' technology.

Arvind Ramnani asked how broader AI advancements enable Upstart to 'leapfrog' and if any previously special internal developments are now broadly available, making Upstart exposed, and also inquired about Upstart's major AI partners for apps and tools.

Answer

Paul Gu, Co-founder and CTO, stated Upstart is grateful for AI progress, not threatened, explaining that general AI excels at human-like tasks, but loan underwriting is a complex math problem requiring specific data and proprietary algorithms built over 12 years. He confirmed they use models from 'all of the foundation model providers' for various use cases, but their core lending models are proprietary and not directly built on others' technology, emphasizing their unique and hard-to-replicate tech stack.

Arvind Ramnani from Piper Sandler sought clarification on the extent to which recent performance was driven by internal model improvements versus macro tailwinds, and asked about the future balance between committed and opportunistic capital.

Answer

CEO Dave Girouard attributed the strong Q3 growth primarily to significant upgrades from their 'Model 18' AI, rather than macroeconomic factors like rate cuts. Regarding funding, he stated that while committed capital is currently over half of the supply and highly valued, the ideal long-term structure includes a healthy mix with opportunistic, 'spot market' capital.

Arvind Ramnani from Piper Sandler sought to understand the broader business strategy behind expanding into subjects like math, music, and chess, and asked about the common thread among these learners.

Answer

CEO Luis von Ahn clarified that all new subjects are monetized through the same subscription model. The strategy is to evolve Duolingo into a broad educational app for subjects that require long-term learning. The common user characteristic is a desire for self-improvement, with users shifting time from mobile games and social media to productive learning.

Arvind Ramnani asked about the long-term relationship between DAU and revenue growth, the performance of this year's Duocon versus last year's, and the strategic distinction between features in the Super and Max tiers.

Answer

CEO Luis von Ahn Arellano and CFO Matt Skaruppa explained that while DAU growth is strong, the company has many levers to drive revenue, so a fixed ratio is not the focus. Von Ahn noted Duocon's live viewership tripled year-over-year. He clarified that currently, Super is for 'convenience' features, while Max is for 'efficacy,' centered on the Video Call with Lily for conversation practice.

Arvind Ramnani asked if the combined Clearwater-Enfusion entity could accelerate the industry's decision-making pace and whether the new solution could create a cascading effect, compelling clients and competitors to adopt it.

Answer

CEO Sandeep Sahai affirmed that the combination enhances the growth rates of both companies by providing a significant new reason for clients to switch. He agreed that the long-term vision of a single platform with a shared data plane could be a 'complete game changer' and create a cascading effect. He emphasized that eliminating the vast industry-wide expense of reconciling data between front, middle, and back-office systems is the core promise, and the two companies' modern, cloud-native architectures uniquely position them to achieve this.

Arvind Ramnani of Piper Sandler asked where the benefits of GenAI are appearing in financial results and if recent industry AI developments alter their strategy. He also questioned why industry growth has moderated and if a return to 20%+ growth is possible.

Answer

Executive Chairman Steven Beauchamp responded that AI's value is embedded throughout the business via improved client experience, higher margins, and product differentiation, rather than as a single monetized feature. President and CEO Toby Williams attributed growth moderation to the law of large numbers but stressed their strategy is to remain the highest-growth player by delivering the most modern platform, balanced with a focus on profitability.

Arvind Ramnani sought clarification on the components of the full-year guidance raise, asked about the potential for further upside, and questioned if AI development would require a step-up in investment.

Answer

CFO Ryan Glenn broke down the recurring revenue raise into two-thirds from Airbase and one-third from organic performance and momentum. He noted a return to a beat-and-raise cadence is the goal. Executive Chairman Steven Beauchamp stated that AI investment is not new and he does not foresee a need for an outsized step-up in R&D spending to fund the AI roadmap.

Arvind Ramnani questioned if Intapp is maintaining flexibility with its AI technology beyond Microsoft and asked about the potential risk of Microsoft increasing costs for its AI services.

Answer

CEO John Hall affirmed that Intapp has its own long-standing AI capabilities and is not technologically exclusive to Microsoft, though the partnership is highly strategic given the Microsoft-centric nature of its end markets. Regarding costs, he acknowledged Microsoft's strong position but believes the broader AI landscape, including open-source options, creates a competitive balance that Intapp is well-positioned to leverage.

Arvind Ramnani requested an update on BETI adoption among the slower-to-adopt legacy customer cohort and asked about progress in the company's international markets.

Answer

CEO Chad Richison stated that while all new clients adopt BETI, the company continues to work with legacy clients to help them see its value and drive adoption. CFO Craig Boelte provided an international update, mentioning a recent win with a manufacturing company with global locations and noting increased adoption from U.S. companies with multinational operations.

Arvind Ramnani requested quantification of AI's revenue and cost benefits and asked how Intuit balances lower-ARPU AI answers with higher-ARPU human expert interactions in TurboTax Live.

Answer

CEO Sasan Goodarzi and CFO Sandeep Aujla addressed the questions. They stated that AI's revenue impact is currently immaterial and not factored into FY25 guidance, though it will be a future growth driver. On the cost side, AI investments are already in the run-rate and are asset-light (via AWS), while also driving internal productivity. Goodarzi corrected the premise of the second question, explaining that AI actually leads to higher ARPU over time by building customer confidence and driving greater attachment of both paid services and human expert help.

Asked about any company-specific drivers of improvement despite a flat macro environment and about the nature and timing of large-scale Gen AI projects.

Answer

Company-specific improvements are driven by increased outbound sales efforts, a focus on resilient verticals, and expanded service offerings like DAMO. Regarding Gen AI, a few large-scale projects are in flight, and the complexity (and project size) increases from using existing LLMs to RAG, to fine-tuning, and finally to building custom LLMs, which are expected to become more common late this year or next.