AI for production — autonomous agents that run, triage, and fix live software, at a $1.5B valuation
HQ: San Francisco, CA (NY and Austin offices) | Founded: 2024 | Employees: ~128 (Mar 2026) | Stage: Late Series A — $1.5B valuation (Apr 2026) | Website: https://resolve.ai
Resolve AI builds 'AI for prod' — multi-agent AI systems that run and operate software in production. Founded in early 2024 in San Francisco by Spiros Xanthos and Mayank Agarwal — who met at UIUC, co-created OpenTelemetry (now a CNCF standard), and together built Log Insight (acquired by VMware) and Omnition (acquired by Splunk, where they led observability as SVP/GM and Chief Architect) — Resolve emerged from stealth on Sept 24, 2024 with a Greylock-led $35M seed and the goal of auto-resolving 80% of alerts. Its agents join every on-call rotation, triage alerts within minutes, run runbook checks, compute blast radius, and recommend actions; a coordinated team of specialized agents investigates multiple hypotheses in parallel with independent verifiers, and engineers steer from a shared Workbench that also executes governed remediations — commit reverts, GitHub Actions, alert silencing. The platform pairs frontier models with domain-specialized, post-trained production models, keeps a continuously-updated queryable graph of services, dependencies, deploys, and team knowledge, and ships 60+ integrations across Slack, Microsoft Teams, Datadog, Grafana, PagerDuty, GitHub, Snowflake, ServiceNow, and more.
Resolve AI is the strongest pure-play bet on the 'operations is the binding constraint' thesis of the AI-coding era: while Cursor and Devin make code cheaper, Resolve makes production safer, with a genuinely distinctive multi-agent architecture (parallel hypotheses + independent verifier agents), a living production knowledge graph, and governed closed-loop remediation used by Coinbase, DoorDash, Salesforce, and Zscaler. Its momentum is real but its economics are extreme — ~$4M ARR (Dec 2025) against a $1.5B valuation implies a story-valued price that any bookings miss could re-rate sharply. The category is crowded from every side (observability incumbents, AIOps, incident management, coding-agent builders), so Resolve's durability rests on domain-specialized models, enterprise workflow depth, and converting its famous founder pedigree into scaled GTM under a newly hired founding CRO.
Browse: Home | Blog | About | All Startups A-Z | View full profile on StartupWiki