Wallarm protects web applications, APIs and AI agents running in production. It is for security and platform teams whose applications expose large numbers of APIs, many of them sitting behind AI applications and agents that call models and tools on their own.
The aim is one control layer in front of all of that traffic, rather than separate tools for web attacks, API abuse and agent activity.
At A Glance
| Vendor | Wallarm |
|---|---|
| Category | Web application and API protection, AI runtime enforcement |
| Identities covered | AI agents, AWS IAM roles, API users and sessions |
| Deployment | Security Edge SaaS, Security Edge in VPC, self-managed nodes, Kubernetes |
| Customers | Panasonic, Samsung, Dropbox, Miro, Rappi |
| Certifications | AWS Security Competency |
| Market classification | Application Security; also Cloud Security |
What Wallarm Does
Security teams rarely have a full list of their APIs, and agent traffic adds model calls and tool invocations that traditional web application firewalls do not see. Wallarm starts from live traffic: it learns what endpoints and agents exist, detects abuse by behaviour rather than signatures, and blocks at session level so a single user or agent can be stopped without cutting off everyone else. The same view is linked to the cloud resources and permissions behind each endpoint. Enforcement can run inside the customer’s own VPC, so traffic does not have to leave their environment.
Products
WAAP
WAAP is a cloud-native web application and API protection platform. It stops injection, XSS, RCE, SSRF and the rest of the OWASP Top 10, along with OWASP API Top 10 attacks such as BOLA and broken authentication. It also covers bots, account takeover, scraping and layer 7 DDoS, and applies virtual patches for 0-day vulnerabilities.
API Security
API Security builds one from live traffic, including shadow, zombie and rogue endpoints, and generates OpenAPI specifications from what it sees. Abuse such as account takeover, scraping and malicious bots is detected by behaviour and blocked at session level rather than by IP address. Teams that already have a specification can upload it for testing.
- Inline protection for REST, GraphQL, gRPC, SOAP and WebSocket
- Spec-based API testing
AI Hypervisor
The AI Hypervisor is the runtime enforcement layer for AI agents in a Kubernetes cluster. It finds every model provider, MCP tool and agent framework running in a cluster, approved or not, including providers such as Anthropic, OpenAI, Bedrock and Vertex, frameworks like LangChain and LlamaIndex, and custom MCP servers.
- Block, rate limit or revoke an agent session live, keyed on JWT subject or trace ID, without restarting pods
- LLM cost broken down by user and team
- Continuous AI-SBOM records and session audit logs, mapped to articles of the EU AI Act
Infrastructure Discovery
Infrastructure Discovery gives cloud and security teams a continuous, cross-account picture of their AWS estate.
- Topology of accounts, regions, VPCs and resources, read through one read-only role per account
- Findings pulled in from AWS security services
- Analysis of over-permissioned IAM roles and what each one can reach
- Drift detection attributed through CloudTrail
- Custom detection and triage rules written in CEL
How It Works
Wallarm runs inline, out of band, as a sidecar or in Kubernetes. Customers choose between three hosting modes: Security Edge SaaS, Security Edge in VPC, and a self-managed edge inside their own VPC.
| Component | What it does |
|---|---|
| Security Edge | A DNS record change routes traffic through Wallarm’s distributed network, which is the quickest way to start. |
| Self-managed nodes | Nodes run as NGINX modules, as a Kubernetes Ingress controller or Envoy sidecar, or alongside Envoy, Kong and MuleSoft, in AWS, GCP, Azure, IBM Cloud or private data centres. |
| Out-of-band mode | eBPF analysis reads traffic without sitting in the request path. |
| AI Hypervisor | The hypervisor installs as a Kubernetes DaemonSet through Helm on EKS or any other cluster, with no sidecars, and uses kernel-level eBPF instrumentation so application code does not change. Infrastructure Discovery reads each AWS account through one read-only role and can mirror traffic out of band through VPC Traffic Mirror or Gateway Load Balancer. |
Identities And Systems Covered
| AI agents | Model calls, tool invocations and outbound connections, each tied to the user who triggered them. |
|---|---|
| Cloud IAM roles | Over-permissioned AWS roles, including roles that AI workloads assume. |
| API users and sessions | Account takeover and broken authentication, stopped for the whole session. |
| Systems | AWS accounts and Kubernetes clusters including EKS. |
Integrations
| Area | Integrations |
|---|---|
| Incident response | PagerDuty, OpsGenie, VictorOps, Jira |
| Security | Splunk, Sumo Logic, QRadar, Demisto |
| Code and CI | GitHub, GitLab, Jenkins, CircleCI |
| Communication | Slack, Microsoft Teams, Twilio, Gmail |
| Observability | Prometheus, Grafana, Datadog |
| AWS security services | GuardDuty, Inspector, Macie, IAM Access Analyzer |
Company Background
| Founded | 2013, with Ivan Novikov as co-founder. |
|---|---|
| Headquarters | Austin, Texas. |
| Leadership | Shayne Higdon (CEO). |
| Investors | Toba Capital, Y Combinator and Partech. |
| Recognition | Included in the GigaOm Radar for API Security. |
Note: This overview is compiled from Wallarm’s publicly available information as of 4 November 2026. It is provided for information only. NHI Mgmt Group does not endorse any vendor, and nothing on this page represents the opinions of NHI Mgmt Group.