Gabriela Borges

Both executives declined to quantify 2027, pointing to durable AI, SASE, OT and regulatory themes and deferring numbers to January or February, so the modelling question went unanswered.

Ken mentioned closer collaboration with leading AI companies on vulnerabilities and automation, plus building proprietary infrastructure with better security, but gave limited specifics.

Ken explained that AI data centers and sovereign SASE/AI converge on the same FortiGate/FortiOS platform, creating compounding demand from a single product architecture.

Ken revealed Fortinet may extend end-of-service instead of forcing upgrades, charging higher service fees as a win-win. Positioned new functionality as the primary growth driver over refresh, citing SASE, ZTNA, internal segmentation, and AI agent security. Christiane confirmed customers are upgrading edge technology alongside SASE adoption.

Ken described a broad infrastructure approach to AI security rather than focusing on a single area. Cited long-term investments in ASIC chips, system-level and infrastructure-level solutions as preparation for the evolving AI landscape.

Christiane expressed confidence in continued product growth driven by upgrade activity, new use cases, new customers, and OT. Ken highlighted FortiOS 8.0 launch and new ASIC next year as specific catalysts.

Christiane confirmed macro resilience and improved sales confidence, supporting the billings guidance raise. Positioned Q1 hesitancy as resolved.

Christiane acknowledged Street expectations may have been elevated while noting refresh tracking difficulty varies by customer size. Ken contextualized refresh as a small percentage of total business given products are 12-15 years old.

Christiane noted positive feedback from Q2 events, good channel activity and programs, but acknowledged PO risk from macro news flow. No actual deal delays observed yet.

Christiana explained that subscription growth requires upselling additional services with new hardware, and Ken added that new hardware has more capacity to run additional functions including unified SaaS.

Seifert cited TAM expansion from $30B at IPO to $300B+, the Act 3/Act 4 flywheel raising deal sizes, and larger cohorts growing faster as reasons growth could re-accelerate.

Prince explained Cloudflare abstracts underlying silicon, matching workloads to optimal resources. Sees Anthropic managed agents as positive, citing Dynamic Workers adoption with one AI studio going from zero to 1 million instances in 15 days.

Seifert confirmed pool of funds is now in its sixth quarter and has become standard in go-to-market. Highest-ever renewal rate validates stickiness and expansion hypothesis.

Prince described incredibly strong pipeline driven by embedding developers directly with customer teams to understand use cases. Customers discover Cloudflare can deliver faster, more scalably, and more cost-effectively than hyperscalers. Not yet at $100M per quarter cadence but confident in getting there.

Prince said primary inference competition remains hyperscalers. When Cloudflare is in the conversation, they demonstrate better TCO and performance. Once customers learn the better way, Cloudflare is very sticky long-term.

Prince explained Cloudflare is not capacity constrained due to giant scheduler architecture that moves workloads to wherever spare capacity exists globally. CPU utilization runs at 70-80%, and GPU utilization is being driven up similarly. Seifert added that off-the-shelf equipment allows deployment that generates revenue before hardware is even paid for.

Prince surprised that AI companies were receptive, understanding content is fuel for their engines. Key focus is ensuring level playing field so no single company gets unique content advantage.

Prince suggested the free tier's content could become more valuable than enterprise revenue if Cloudflare helps content creators monetize AI access. Strategic value over extraction.

Prince described bottoms-up self-serve approach for initial adoption, then strategic targeting of specific accounts. Ali Cabral leading a developer speedboat for targeted enterprise pursuit.

George walked through the identity control plane for non-human identities, data protection, runtime visibility and exposure management as the stack a customer needs to move agents into production safely.

George pointed to rising next-gen SIEM adoption and cloud workloads, attributing it to the data gravity and disruptive SIEM pricing of the platform.

Kurtz acknowledged LLMs are great at data sorting and aggregation but cannot stop breaches in real time. CrowdStrike uses bespoke models with vertical expertise plus frontier models. Better together approach with agents talking to agents.

Podbere argued Flex drives continuous net new ARR as customers consolidate further, emphasizing bigger/longer deals and lower TCO as structural benefits not cyclical ones.

Kurtz framed EDR as foundational telemetry enabling all other modules, noting the service layer (Overwatch, Complete) plus Charlotte AI creates a winning combination competitors cannot match.

Kurtz explained Flex customers fund expansion by replacing point products with ROI-backed consolidation; larger commitments unlock deeper discounts and remove procurement friction.

Podbere explained CCP expiration creates natural upsell opportunity since customers are already using and liking the products, driving conviction in back-half reacceleration.

Burt cited slightly larger deals, $1.3B total deal value, 9+ module adoption on average, and multiples larger than typical contract value as evidence supporting back-half FY2026 reacceleration.

Praveer said guidance waits for sustained trends; beta products don't yet draw paid credits, leaving upside and driving investment.

Dylan expects more tools, some complementary and some competitive, with design as the layer above commoditized code.

Dylan said responsibilities, not roles, are blurring as non-designers engage in design tasks; Praveer noted broader IT budgets.

Dylan emphasized Make-plus-Design usage and the code-design round-trip as the platform differentiator.

Praveer explained Make billing infrastructure is being built and credits are not yet strictly enforced for full seats; no monetization figures given.

Brian said R&D rising from 5.5% in 2024 to ~8-9% in 2026 with ~$20M from COGS geography migration. Remainder from AI and product innovation investments across Tyler. Expects 2026 level to be closer to long-term steady state.

Lynn said courts document automation and ERP priority-based budgeting seeing most AI receptiveness. Workforce aging driving demand for digital workers. Brian added trust factor critical - clients prefer Tyler over point solutions or startups.

Brian said ~25% annual flip growth through peak in 2027-2028, with dollar value increasing more than number as large court clients flip, maintaining ~1.7x maintenance-to-SaaS uplift.

Lynn cited direct client feedback as basis for confidence, with Arizona client example where DOGE-driven cuts actually increase automation demand for Tyler products.

Brian said commoditized payments exposure is minimal beyond Texas, only Florida is payments-only but with differentiated services. Lynn noted executive comp excludes merchant fees.

Olivier said every sale must make or save customers money, that Datadog increasingly helps rein in AI costs, and that Infinite Cardinality Metrics addresses a longstanding source of unpredictable-bill frustration.

Olivier explained training is transitioning from artisanal research to production workload at scale. David added 6,500 customers using AI integrations representing 80% of ARR, with training contributions still early.

Pomel emphasized customers must see cost savings or revenue gains for every product adopted. Using Datadog's platform rather than adding another vendor typically costs less.

Pomel distinguished between post-hoc LLM analysis (where general tools can compete) and in-stream real-time analysis across orders of magnitude more data, where Datadog's data plane embedding is uniquely positioned.