A panel discussion on artificial intelligence and enterprise software focused on how AI is likely to reshape—rather than rapidly replace—large, mission-critical platforms. Speakers argued that investor and executive anxiety has outpaced reality, particularly around the idea that large language models (LLMs) could quickly “vibe code” replacements for complex ERP and other transactional systems.
Rimini Street and Zeta outline scale and customer focus
Seth Ravin, CEO of Rimini Street (NASDAQ:RMNI), said the company serves thousands of customers globally, including governments and the military, and primarily works with large organizations—often those with $10 billion or more in annual revenue. Ravin said Rimini Street focuses on reducing customers’ total cost to serve across labor, systems, and processes, while increasingly layering “agentic AI solutions” on top of existing environments.
David Steinberg, CEO of Zeta, described Zeta as a marketing automation platform that uses first-party data and proprietary AI algorithms that the company began developing in 2017. Steinberg said Zeta has 603 clients globally and stated that “51% of the Fortune 100” are clients. He cited a Forrester-reported “600% return on investment” for spending through Zeta’s platform.
On company size and financial profile, Steinberg said Zeta expects $1.755 billion in revenue at the midpoint of its range for the year, nearly $400 million in EBITDA, and that 60%+ of EBITDA should convert to free cash flow. He also said the company is on track for a fourth consecutive year of 30%+ compounded top-line growth, with faster growth in EBITDA and free cash flow over that period.
Panelists push back on the idea that AI will quickly replace core platforms
Ravin argued that replacing major ERP platforms is not a near-term proposition due to the scale and compliance requirements involved, citing “27,000 processes,” “millions and millions of lines of code,” and the need to meet auditing and public company standards. He said changes will be gradual, noting that Rimini Street was founded to keep large enterprise systems running longer than vendors typically prefer, with customers often seeking stability over frequent upgrades.
Ravin added that while he believes “ERP software…is dead” in the long run and will be replaced by a new paradigm, that shift is likely to take time. He characterized near-term expectations of rapid replacement as “overanxiety.”
Steinberg compared current AI fears to earlier technology cycles involving the internet, mobile, and cloud computing, arguing that companies adopting those technologies often strengthened their positions. He said “thin applications” and workflow tools without proprietary data or deep integration are more likely to struggle, while companies that create measurable intelligence and ROI could thrive by embracing AI.
Moats: proprietary data, integration, governance, and trust
Steinberg emphasized Zeta’s proprietary data and infrastructure as key defenses. He said Zeta’s data cloud includes 552 million active people and described its technical requirements in terms of large-scale ingestion and computation to predict intent, such as whether an individual might churn from a wireless provider or seek a new credit card.
He also stressed Zeta’s approach to privacy and data handling, saying the company never sells its data and does not share personally identifiable information with clients, using a “Zeta ID” instead. Steinberg said Zeta does not feed its data into LLMs and relies on proprietary models, adding that clients have built “thousands of AI agents” using Zeta’s AI agent studio. He cited a 120% net retention rate as evidence of stickiness.
Ravin said organizations face an overwhelming amount of “AI washing,” which he believes is contributing to CIO and CFO indecision. He argued that many businesses could see major gains through process improvement and automation even before adding AI, and he described Rimini Street’s use of checklists and frameworks to help customers identify where AI agents can generate value.
Both speakers discussed the complexity of enterprise adoption, including security concerns, procurement processes, and audit requirements. Ravin also highlighted the challenge of AI pricing models, comparing today’s complexity to early long-distance telephone plans that eventually evolved into unlimited pricing.
Partnerships and product initiatives: OpenAI and ServiceNow
Steinberg discussed Zeta’s partnership with OpenAI, describing it as focused on building “Athena,” a voice-enabled “super agent” designed to help clients navigate Zeta’s user interface and execute tasks via voice commands. He said Zeta has a joint engineering team working with OpenAI and positioned “applied AI” and smaller models layered on top of foundational LLMs as the likely next wave. Steinberg said Zeta is not exposing its proprietary data to OpenAI and is not using OpenAI’s foundational models for intent scoring.
Ravin discussed Rimini Street’s partnership with ServiceNow, saying ServiceNow and others offer AI toolkits but customers want complete solutions rather than tools. He said the partnership aims to bring “agentic” capabilities to ERP and transactional environments to reduce labor costs and increase flexibility. Ravin described deploying “Band-Aids” on top of existing ERP systems—20 initial solutions—aimed at addressing real problems and shortening deployment timelines from months to weeks.
Internal productivity and who is most at risk
On where AI may pose the greatest competitive threat, both panelists pointed to less complex, standalone workflow tools. Steinberg said such products can be replicated quickly, particularly when they lack proprietary data and deep integration. Ravin framed the risk using a “sliding scale of complexity and mission criticality,” saying low-end tools are “just gonna get killed,” while high-complexity, mission-critical systems are better positioned to use AI to improve operations.
Steinberg said Zeta expects to grow 30%+ while ending the year with fewer employees than it started with, citing internal adoption of tools from Microsoft, Anthropic (Claude), Gemini, and others, and said engineering productivity has increased 125% over the past 12 months. Ravin said Rimini Street will keep “humans in the loop” due to mission-critical responsibilities, but described using AI to improve engineer efficiency, accelerate response times, and reduce time-to-resolution by 28%, including achieving a “70 seconds” response time with an engineer globally, 24/7.
On governance, Steinberg cautioned against appointing a single “AI czar,” favoring broad organizational adoption supported by education, incentives, and measurement. He also emphasized the importance of making employees feel safe using AI so they do not view adoption as a threat to their jobs. Ravin said he recommends a separate innovation team outside IT to enable experimentation and “fail fast,” with successful projects later adopted more broadly.
In closing, the discussion framed AI as a potentially significant tailwind for “good software companies,” particularly those with deep integration, defensible data, and clear ROI, while noting that meaningful platform replacement is likely a multi-year journey rather than an imminent disruption.
About Rimini Street (NASDAQ:RMNI)
Rimini Street, Inc (NASDAQ: RMNI) is a provider of enterprise software support services, specializing in third-party maintenance for mission-critical applications from leading technology vendors. The company offers comprehensive support for ERP, CRM and database environments, with coverage for systems from providers such as Oracle and SAP. Through its proactive system monitoring, performance tuning, regulatory and tax update services, Rimini Street aims to extend the lifecycle of enterprise applications while delivering service levels comparable to or exceeding those of original software vendors.
Founded in 2005 by technology entrepreneur Seth Ravin, Rimini Street has grown from a startup into a publicly traded company following its initial public offering in March 2018.
