We've been investing in AI-native enterprise software since 2020, and the pattern of this technology cycle is becoming clearer than it has been at any prior point. Every technology cycle of this magnitude has three phases: experimentation (the capability becomes available, teams try everything, many things don't work), rationalization (teams learn what actually works and what doesn't, the bad experiments get shut down, standards begin to emerge), and maturity (the patterns that work become standard practice, the infrastructure that supports those patterns becomes ubiquitous, and value creation shifts from the applications layer to the infrastructure layer that makes reliable operation at scale possible).
Enterprise AI has been in the experimentation and rationalization phases simultaneously for the past two years. The experimentation phase was characterized by broad enthusiasm, POC proliferation, and inflated expectations — companies tried AI in dozens of workflows and most of those efforts stalled at production. The rationalization phase is underway now: the workflows where AI genuinely delivers reliable value at scale are becoming clear, the procurement and governance frameworks for enterprise AI are stabilizing, and the teams building AI applications have developed a more accurate model of what it actually takes to deploy and maintain AI in production.
What maturity looks like
The maturity phase, which we believe is beginning now, is defined by several characteristics. AI-native workflows become standard operating procedure in specific high-value enterprise functions: AP automation, contract analysis, customer service triage, code review, regulatory compliance monitoring. The infrastructure stack underneath those workflows — orchestration, observability, governance, data access, human-in-the-loop handoffs — gets standardized and productized. Organizations build internal AI operations functions to manage that infrastructure with the same rigor they apply to any other critical enterprise system. The variance in outcomes between organizations that have invested in AI infrastructure and those that haven't becomes large enough to be measurable in operational metrics.
For investors, the maturity phase is when the most durable companies in a technology cycle are built. The experimentation phase produces many funded companies, most of which don't survive the rationalization. The maturity phase rewards companies that got the infrastructure right early, that have the operational credibility to win enterprise trust, and that built for the requirements of production systems rather than the requirements of demos.
Looking at the Ridgepoint portfolio, the companies that are performing best are the ones that treated production reliability as a design constraint from day one, not a future milestone. Inngest was built around the assumption that workflows would fail and that failure needed to be handled gracefully — before most AI orchestration products had that design philosophy. Portkey was built around the assumption that AI systems needed semantic observability, not just infrastructure monitoring — before enterprise AI teams fully understood why that distinction mattered. The pattern is consistent: the companies we're most confident about are the ones that understood the production requirements of the maturity phase before the rest of the market did.
We're entering the most interesting phase of this technology cycle. The groundwork has been laid. The workflows have been identified. The governance frameworks are maturing. The infrastructure investments that enable reliable, scalable, governed AI-native workflows are the investments that will define the enterprise software landscape for the next decade. We remain deeply focused on finding and backing the teams doing that infrastructure work with genuine technical depth and long-term architectural vision.