Resolve AI

What Resolve AI does

Resolve AI provides AI agents that run software for engineering teams, handling on-call, incidents, and daily production work. This allows engineers to focus on building new features. The platform combines expertise across teams, operates various tools, and captures tribal knowledge of unique systems. It drives faster incident investigations, gets to root cause for complex issues, and automates operational tasks. Resolve AI offers pre-built agents and allows users to build their own, integrating with existing ecosystems via MCP, API, and Skills. The platform is designed with enterprise security and production readiness in mind, including SAML SSO, RBAC, data protection, customer isolation, and auditable activity, with SOC 2 Type II, GDPR, and HIPAA compliance.

Products

• Resolve AI AgentsAI agents that drive on-call, incidents, and daily operational tasks in production. Engineers step in to direct and take action. Users can use Resolve's agents, build their own, or both.
• On-call agentDelegates on-call responsibilities to agents, who participate in every on-call rotation to triage and investigate alerts, post findings before engineers are paged, silence noise, and route to the right team.
• Incidents agentGets engineers to root cause by having agent teams investigate incidents in parallel across code, infrastructure, and telemetry. Engineers steer through Workbench to remediate.
• Background agentAutomates operational tasks such as deployment monitoring, operational reports, and resource optimization, running on a schedule or triggered by events.
• Custom agentAllows users to build their own agents using Resolve AI capabilities exposed as MCP, API, and skills, enabling them to call Resolve for production context, investigation, and remediation.
• Resolve AI PlatformThe underlying platform that powers all agents, pairing frontier models with domain-specialized models, building a queryable graph of services, governing actions, and providing 60+ pre-built integrations.
• WorkbenchAn interface for engineers to co-work with agents during incidents, interrogate findings, evidence, or theories, and steer investigations.

Key capabilities

• AI agents for on-call, incidents, and operational tasks
• Root Cause Analysis (RCA) for complex issues
• Ability to build custom agents with MCP, API, and Skills
• Delegation of on-call to agents for triage and investigation
• Co-working with agents for incident resolution
• Automation of operational tasks with background agents
• Up to 5x faster MTTR (Mean Time To Resolution)
• 75% higher productivity
• Enterprise-grade security: SAML SSO, RBAC, admin controls
• Data protection: redaction, encryption, retention controls
• Customer isolated data
• Auditable activity and support access logs
• No external training of models with customer data
• Vulnerability SLAs
• SOC 2 Type II certification
• GDPR compliance
• HIPAA compliance
• 60+ integrations across code, infrastructure, telemetry, knowledge, and collaboration tools
• Pairs frontier models with domain-specialized models
• Builds a queryable graph of services, dependencies, deploys, and team knowledge
• Governed actions within guardrails (silencing alerts, reverting commits, opening PRs, executing GitHub workflows)
• Flexible and secure integrations with MCP, APIs, and Webhooks
• BYO Custom tooling integration
• Minimal data access with full control (read-only, least privilege, no data ingestion)
• Customer data used only to train their models (no data mixing, no cross-customer models, exclusive fine-tuning)
• Unified Enterprise Control with secure gateway (Resolve AI satellite)
• Granular control over data (query specific data, customize metadata scraping frequency, limit access)
• Streamlined and secure authentication (SSO, RBAC, service accounts, secure tokens)
• On-call rotation participation by agents
• Alert correlation across observability stack
• Severity assessment and blast radius identification with evidence
• Alert resolution without changing context (silencing noise, executing GitHub Actions, routing to teams)
• Findings, priority lists, and actions surfaced in collaboration tools (Slack, MS Teams, CLI)
• Specialized agent teams for harder investigations (The Council: Lead, Triager, Investigator, Verifier, Mitigator)
• Evidence Trail for root cause findings
• Parallel hypothesis pursuit during investigations
• Ability to redirect agents and add context
• Verified root cause with production evidence
• Automated task execution on schedule or trigger
• Template library for operational tasks (deploy health monitor, daily digest, alert triager, resource report)
• Curated priority list with pre-investigated items
• Chat interface for live debugging and multi-turn investigations
• Continuous monitoring of Kubernetes clusters
• Pod failure investigation
• Deployment change correlation
• Node and resource pressure awareness
• Automatic service dependency mapping
• Workload health as investigation context
• Cross-stack correlation for Kubernetes data
• Scoped read-only access for Kubernetes
• Deployment event tracking for Kubernetes
• Pod lifecycle monitoring for Kubernetes
• Node condition ingestion for Kubernetes
• Namespace-scoped deployment for Kubernetes
• Cloud log analysis for Google Cloud
• Cross-project investigation for Google Cloud
• Deployment and change correlation for Google Cloud
• Secure, credential-flexible access for Google Cloud (Service Account Key, Workload Identity Federation)
• Multi-scope support for Google Cloud logs
• Sensitive data redaction at the Satellite for Kubernetes and Google Cloud

Use cases

• Faster incident investigations
• Improved on-call experience
• Root cause analysis for complex issues
• Automation of daily production work
• Triaging and investigating alerts
• Resolving incidents with co-working agents
• Automating operational tasks (deployment monitoring, operational reports, resource optimization)
• Reducing MTTR (Mean Time To Resolution)
• Increasing engineering productivity
• Ensuring enterprise security and compliance
• Integrating with existing production ecosystems
• Debugging production issues
• Optimizing costs in financial systems
• Coding with production context
• Reducing dashboard hopping for NOC and SRE teams
• Monitoring SLO health
• Generating daily 'what breached' reports
• Understanding recent changes around a service
• Correlating signals across observability stack
• Distinguishing genuine issues from transient spikes in alerts
• Tracing failures to their origin across metrics, APM traces, and logs
• Detecting deployment impact
• Load test and scheduled event awareness
• Investigating pod failures in Kubernetes
• Correlating deployment events with telemetry in Kubernetes
• Monitoring node conditions and affected workloads in Kubernetes
• Mapping service dependencies in real-time in Kubernetes
• Analyzing cloud logs for error patterns and tracing failures in Google Cloud

AI approach

Resolve AI develops AI agents that automate on-call, incident response, and operational tasks for software engineering teams. The platform uses a combination of frontier models and domain-specialized models, post-trained on production data, to identify root causes, triage alerts, and automate workflows. It builds a queryable graph of services, dependencies, and team knowledge, and can operate tools within guardrails.

Tech named: frontier models, domain-specialized models, Google DeepResearch, Gemini Agents, Meta superintelligence research

Industries served

• Software Development
• Financial Systems
• E-commerce
• Zero-Trust Network Security
• Cloud Computing
• CRM

What it says sets it apart

• AI agents that run software, not just assist
• Combines expertise across teams, operates all tools, and captures tribal knowledge
• Gets to root cause in the most demanding production environments
• Allows building custom agents with MCP, API, and Skills
• Drives up to 5x faster MTTR and 75% higher productivity
• Designed for stringent compliance standards (SOC 2 Type II, GDPR, HIPAA)
• No external training of models with customer data; customer data is isolated and used only for their models
• Pairs frontier models with domain-specialized models post-trained on production
• Builds a continuously updating queryable graph of services, dependencies, deploys, and team knowledge
• Governed actions within guardrails for automated remediation
• 60+ pre-built integrations with automatic handling of schema changes, auth rotations, and rate-limits
• Minimal data access with full control (read-only, least privilege, no data ingestion)
• Secure gateway (Resolve AI satellite) for flexible data access and control
• Granular control over data querying and access
• Autonomous investigations that identify underlying causes hours before human intervention
• Provides clear working theories with production evidence, not just walls of telemetry
• Co-creators of OpenTelemetry on the founding team, bringing deep observability expertise
• AI Researchers from leading labs (Google DeepResearch, Gemini Agents, Meta superintelligence research) on the team
• Domain experts in production systems from Google, Meta, Microsoft, Databricks
• Backed by AI pioneers and builders of Enterprise Software (Jeff Dean, Fei-Fei Li, Reid Hoffman, Amjad Masad, Srinivas Narayanan, Thomas Dohmke, Matt Garman, Jeff Lawson, Sridhar Ramaswamy, Eric Glyman)

Funding rounds we track

From the AI funding tracker — rounds as reported by the linked publications.