Resolve AI, a startup created by former executives at Splunk, has become the latest $1 billion headline valuation from a Series A fundraising led by Lightspeed Venture Partners, according to people familiar with the deal. The company creates an autonomous site reliability engineer—software to detect, diagnose, and fix production issues in real time—meant to reduce outages without depending on always-on humans.
Deep Observability DNA Brought by Ex-Splunk Founders
Resolve AI is led by CEO Spiros Xanthos and CTO Mayank Agarwal, who previously worked together at Splunk, where Agarwal was the chief architect for observability. The two were also co-founders of the distributed tracing startup Omnition, which Splunk acquired in 2019, and their partnership dates back to graduate work at the University of Illinois Urbana-Champaign.
The company came into being less than two years ago and quickly attracted powerful backers. It has raised a $35 million seed round, led by Greylock, with investments from legendary AI leaders Fei-Fei Li and Jeff Dean. Resolve AI and Lightspeed declined to comment on the latest financing.
Autonomous SREs of the Future: Reducing Downtime
With companies spanning frameworks, Kubernetes, and multi-cloud estates, keeping systems running has become a 24/7 firefight in the digital age. The platform on which Resolve AI is built assumes the conventional duties of a site reliability engineer in monitoring signals, correlating anomalies, determining root cause, and applying fixes without waiting for human runbook steps. Think of it as a melding of observability, incident response, and remediation in a closed loop.
The stakes are high. Uptime Institute estimates in its annual outage analysis that more than 50% of significant outages now exceed $100,000 in costs—and the most serious ones pass into the millions. DORA’s research demonstrates that it is the high performers who are cutting mean time to recover and change failure rates, but a talent bottleneck makes it difficult to clear that bar. An AI-driven SRE that lowers the noise around alerts and automates safe rollbacks or configuration fixes might have a dramatic effect on both MTTR and OpEx.
Two people familiar with the business—who would not speak for attribution because they weren’t authorized to discuss the company’s business in detail—estimate that Resolve AI has annual recurring revenue (ARR) of around $4 million, early but significant for a product that must integrate deeply into existing observability, incident management, and deployment stacks. The company’s go-to-market is probably design partners in finance, SaaS, and consumer platforms where reliability has existential value.
A Unicorn Headline With Tranched Terms in Financing
Though the round was a nominal $1 billion headliner, it was multi-tranched, according to people familiar with the structure: some of the equity had been bought at unicorn pricing and some at a lower valuation—a blended figure below that $1 billion headline. These types of structures have become common in competitive AI financings, enabling a company to secure a milestone valuation while investors package risk and add capital against progress.
For founders, the move can lock in brand credibility and position the company as a recruiting tailwind; for investors, it bakes in performance gates for things like model accuracy, false positive rates, or customer expansion. Those gates, in reliability software, are often quantifiable outcomes:
- % reduction in alerts
- MTTR improvement
- Percentage of incidents fully auto-resolved
Crowded Field, With Incumbents Watching
Resolve AI is not the only company pursuing autonomous operations. Another AI SRE startup, Traversal, recently raised a $48 million Series A from Kleiner Perkins and Sequoia. Indeed, incumbents in observability and AIOps like Datadog, Dynatrace, New Relic, ServiceNow, and Splunk are layering generative capabilities onto alert triage, runbooks, and remediation. The distinction will rest on safe autonomy at scale: can a system confidently make changes across thousands of services without causing a larger incident?
That poses practical questions Resolve AI will have to answer:
- Great guardrails, thorough approval flows, and rollback assurance
- Deep integrations with incident tools and CI/CD
- Proof that the platform reduces toil without dragging teams under with configuration debt
Referenceability—demonstrating that they reduced on-call pages by 30–50% and shaved double digits off MTTR—is what will move the market.
What Comes Next for Resolve AI and Autonomous Ops
With new money in the bank and increasing interest in autonomous ops, I expect Resolve AI is concentrating on scaling customer success, expanding cloud and Kubernetes coverage (since they are perhaps the only things growing even faster than public service outages), and hardening remediation policies for common failure modes like noisy deployments, resource contention, or misconfigured feature flags.
If the company can convert early ARR into enduring enterprise rollouts and post credible reliability wins, the unicorn headline won’t just be optics—it will be table stakes in a war over whose software is going to automate the concept, and reality, of reliability itself.
