Resolve AI lança SRE-1, um modelo de 27B para análise automatizada da causa raiz de incidentes de produção
O novo modelo SRE-1 e o serviço Primary On‑call alegam até 95% de precisão na causa raiz e lidam com 80% dos alertas sem escalonamento humano.
The model, created with Snowflake’s Cortex Training service, improves single‑investigation pass rate from 6.9% to 38.8% and matches frontier models at lower cost in tests on 49 live production cases.
Primary On‑call, the accompanying service, runs the model on every production alert, investigating telemetry, infrastructure, code and operational knowledge, and can close, mitigate or escalate alerts.
Resolve AI reports that up to 80% of paged alerts are resolved without escalation, mean time to resolution improves tenfold, and root‑cause accuracy reaches up to 95% according to early customers like Salesforce.
The system is currently available only through Resolve AI’s platform; the company has raised over $190 million, including a $40 million Series A extension that valued it at $1.5 billion.
O que isso significa para você
For organizations, the model promises faster, more accurate incident handling, reducing on‑call fatigue and operational downtime. For users of free or locally run models, the launch highlights the growing gap between specialized, high‑cost production AI and community‑available models, underscoring the need for continued investment in open‑weight research to narrow that divide.
Fontes
- Resolve AI's SRE-1 model finds root causes of production alerts as AI agents go on call, SiliconANGLE, 19 Nov 2026
- Agents on the critical path is the only viable way forward to keep up with software change, resolve.ai (official), 18 Nov 2026
Escrito automaticamente a partir das fontes acima, depois verificado pelo nosso código: todo número e nome deve aparecer nelas, e nenhuma frase pode copiar uma fonte. Encontrou um erro? Reporte‑o.
Qual IA você pode usar grátis hoje?
Todo modelo de IA gratuito, classificado por votos humanos cegos, com a memória e velocidade que ele precisa na sua própria GPU.