Why Backflipt? | Runtime Governance for AI Agents, Agent-Ready Enterprise Applications

Why Backflipt?

Answers about the company behind the Lumen platform: runtime governance for AI agents at two decision points, intent and credentials, and agent-ready enterprise applications from the APIs enterprises already publish.

Founder-funded since 2013 · Fortune 500 production deployments · In your VPC

Discover

Every AI agent and every enterprise resource it can reach, each with a risk score.

Trust

Cryptographic agent identity, SPIFFE SVIDs, mTLS, and federation with enterprise IAM.

Govern

Runtime authorization at two decision points: intent at the Gateway, credentials at the Broker.

Observe

Every approval and denial logged with the policy that decided it. Governed transfers watched by Galileo AI.

Two decision points, one platform

Agentic AI fails safe when authorization happens where the traffic is. Lumen places a policy decision and enforcement point in front of intent and another in front of credentials, so agents are governed without re-coding a single one.

Enterprise IAM and identity providers

Bearer tokens · Entra agent blueprints · RFC 8693 token exchange

AI agents

Platform, custom, and shadow agents

No token · JIT token · bearer

Lumen A2A Gateway

PDP + PEP for intent

Smart Policies per PeerGroup

Inline inspection (ICAP RFC 3507)

SPIFFE SVIDs · mTLS

Discovery with risk scores

Lumen POET

Credential Broker

PDP + PEP for credentials

Verify caller with IAM

Exchange for vault-held NHI

OAuth · API key · Basic Auth

Enterprise applications

MCP servers and A2A agents

PeerGroup segmentation

External PAM vault

Every approval and denial logged with the PeerGroup requirement and the evaluating Smart Policy. Agents never hold permanent secrets.

The Lumen A2A Gateway decides intent. The Lumen POET Credential Broker decides credentials. Enterprise IAM stays the source of truth, and standing secrets live only in the vault.

A track record, not a pitch

How long has Backflipt been in business?

Backflipt (Xenovus, Inc.) was founded in 2013 and has operated as a founder-funded, self-sustained business for more than twelve years. The company has production deployments in regulated Fortune 500 environments and has never depended on outside capital to serve its customers.

Who are Backflipt's customers?

Fortune 500 enterprises in regulated industries: financial services, insurance, aerospace and airlines, healthcare, government, and software. Customers include New York Life, Boeing, Express Scripts, the State of California, IDC, and several others. These are environments where audit evidence, data sovereignty, and change control are conditions of doing business, and they shaped how every Backflipt product is built.

Who is behind Backflipt?

The founding team consists of Silicon Valley entrepreneurs with a 25-year record of building enterprise infrastructure companies, multiple successful exits, and granted patents. The team's identity management and single sign-on technology became part of Citrix, and its high-performance data center connectivity technology became part of Broadcom.

That history matters here: the people who built enterprise IAM appliances two decades ago are now applying the same discipline to non-human identity and runtime authorization for AI agents.

What certifications does Backflipt maintain?

Backflipt annually renews SOC 2, ISO, and GDPR certifications, meeting the security and compliance standards regulated enterprises require of their vendors.

What did Backflipt build before Lumen?

A no-code workflow automation platform, in production with Fortune 500 enterprises for over a decade. The platform is built in three layers: a rich iPaaS layer with hundreds of application connectors, a workflow orchestration layer, and a drag-and-drop application layer, all integrated with enterprise IAM systems. Financial services, healthcare, and government organizations use it to integrate enterprise applications and surface the activity insights their business units need. These configurations deploy in customer hyperscaler environments and captive data centers, on a yearly subscription model.

The same platform hosts Backflipt's managed file transfer products: the TransferIQ line in the Products section, including TransferIQ Orchestrate, TransferIQ Orchestrate for AWS MFT, TransferIQ Continuous Touch, and Backflipt Apps for Axway MFT/B2Bi, delivered as the Managed File Transfer Solutions offering. Backflipt rolled out the control plane products for managed file transfer, with the data plane built on the AWS Transfer Family of products.

Lumen is built on that foundation. The connector depth, the application knowledge, and the discipline earned across those regulated deployments extended the Backflipt Platform into the Lumen products for the AI market. Both the workflow automation platform and the Lumen products have gone through enterprise architecture review boards at Fortune 500 companies.

What Lumen is

One platform, four products, a single job: let enterprises scale AI agents safely on the infrastructure they already run.

What is the Lumen platform?

Lumen governs AI agents at runtime. It discovers every agent and every enterprise resource an agent can reach, assigns risk scores to both, and enforces authorization at two decision points:

Alongside governance, Lumen POET makes enterprise applications agent-ready by generating precise MCP servers and A2A agents from existing API specifications, and Lumen AI TransferView watches governed file transfers with Galileo AI. Everything runs in the customer VPC.

What does the Lumen A2A Gateway do?

It sits inline on agent-to-agent and MCP traffic and evaluates Smart Policies per PeerGroup, microsegmentation at the agent layer: only approved agents reach the resources assigned to their group. Inline inspection runs over ICAP (RFC 3507). Agent identity is established with SPIFFE SVIDs or built-in cryptographic attestation plus mTLS, and the Gateway federates with enterprise IAM through bearer tokens, Microsoft Entra agent blueprints, and RFC 8693 token exchange.

Policy grounding follows the OWASP agentic threat model: intent breaking, tool misuse, and goal drift are first-class detections, and every approval and denial is logged with the PeerGroup requirement and the evaluating Smart Policy.

What does Lumen POET do?

APIs were never designed for AI. Lumen POET reads the OpenAPI, Swagger, and RAML specifications enterprises already publish, curates the schema for AI consumption, and generates precise MCP servers and orchestrated A2A agent collections with out-of-the-box prompts, no code required. Generated servers deploy with Helm and Terraform in the customer VPC and coexist with the API management stack already in place, including Kong, Apigee, AWS API Gateway, and MuleSoft.

What does the Lumen POET Credential Broker do?

The Credential Broker is an add-on service that front-ends generated MCP servers and A2A agents. The application admin defines which user, role, or AI agent reaches which features and scope. An agent arrives with no token, a just-in-time token, or a user bearer token; the broker verifies the caller with enterprise IAM and exchanges the token for a vault-held non-human identity (NHI) credential: OAuth, API key, or Basic Auth.

Standing secrets live only in the vault. Agents hold nothing worth stealing.

What does Lumen AI TransferView do?

TransferView turns the events a managed file transfer platform already emits into answers, alerts, and incidents. It watches transfers across Axway MFT, Axway B2Bi, and AWS Transfer Family, applies SLA rules, opens ServiceNow tickets automatically, and puts Galileo AI on the data: it analyzes, suggests, and acts. Enterprises find out about a breached SLA before the partner does, with no change to the MFT platform or the transfer flows.

How trust is established and kept

How do AI agents authenticate?

With workload identity, not shared secrets. The Lumen A2A Gateway authenticates agents using SPIFFE SVIDs or built-in cryptographic attestation, plus mutual TLS on every connection. Identity anchors to the enterprise identity provider through federated IAM: bearer tokens, Microsoft Entra agent blueprints, and RFC 8693 token exchange. This is workload IAM applied to the agent layer, machine identity treated with the same rigor as human identity.

How does Lumen eliminate standing privileges?

Agents never hold permanent secrets. The Credential Broker exchanges verified caller tokens for ephemeral, scope-limited credentials held in an external PAM vault, and passwords and certificates never leave that vault. The result is zero standing privileges at the agent layer, which removes the most attractive target in an agentic estate: long-lived non-human identity credentials.

What audit evidence does Lumen produce?

Per transaction: the caller, the schema decision, the credential type, and the credential lifetime at the Broker; the applied PeerGroup requirement and the evaluating Smart Policy at the Gateway. Interaction logs and PeerGroup access logs are first-class, so compliance teams answer "which agent touched what, and under which policy" from the record, not from reconstruction.

How does Lumen address shadow AI?

Discovery runs continuously: Lumen finds AI agents and the enterprise resources they call, including agents nobody registered, and ranks both by risk. Unknown agents are visible before they are governable, and governable before they can do damage. PeerGroup segmentation then ensures that even an approved agent reaches only the resources assigned to its group.

Which threats does the governance model cover?

Smart Policies are grounded in the OWASP agentic threat model. The Gateway watches for intent breaking, tool misuse, and goal drift on live traffic, while the Broker removes credential theft and privilege escalation paths by keeping standing secrets out of agent hands. Together the two decision points cover both halves of the agentic attack surface: what an agent does, and what an agent holds.

On the infrastructure enterprises already run

Where does Lumen deploy?

In the customer VPC, with Helm and Terraform, on Kubernetes or Docker. On-premises and air-gapped deployments are fully supported. All processing and storage remain under customer control, with no external egress: enterprise data never leaves the environment the enterprise operates.

Does Lumen replace the existing API gateway?

No. Lumen POET coexists with Kong, Apigee, AWS API Gateway, and MuleSoft: the gateway keeps managing API traffic while POET generates and operates the agent surface. If the gateway team is exploring AI gateway, MCP gateway, or MCP proxy features, POET fits right beside them. Deployment is additive, with no forklift upgrades.

Does Lumen work with existing agent platforms?

Yes. Lumen is one vendor-agnostic control plane over agents from the platforms teams already use, including Amazon Bedrock AgentCore, Azure AI Foundry Agent Service, Google Vertex AI Agent Engine, Salesforce Agentforce, ServiceNow AI Agents, and Snowflake Cortex Agents. Enterprises gain visibility and control over agents wherever they run.

Does Lumen work with existing identity and vault investments?

That is the design center. The Gateway federates with the identity provider already in place, the Broker verifies callers against enterprise IAM and draws credentials from the external PAM vault already deployed, and SAML SSO governs administrative access. Enterprises scale AI on the identity and security investments they already have.

Built with the security stack, not beside it

What is Trust No Agent?

Trust No Agent is a joint methodology with IBM and Fortinet for governing AI agents in the enterprise. Fortinet secures the network fabric the agents traverse, the Lumen A2A Gateway and Lumen POET Credential Broker enforce runtime authorization for intent and credentials, and IBM Consulting delivers the methodology at enterprise scale. Backflipt is a Fortinet Fabric-Ready validated technology alliance partner. Enterprises get a deployable, three-party answer to the question every board is now asking: how do we let agents work without letting them roam.

Does Backflipt work with other technology partners?

Yes. Backflipt maintains technology alliances across the security and identity ecosystem and validates integrations with the network, IAM, and PAM platforms regulated enterprises already operate. The partnership model is deliberate: Lumen complements the deployed stack rather than replacing it, which is why the platform carries no requirement to change vendors to adopt it.

Does Backflipt work with system integrators?

Yes. Backflipt engages leading system integrators and channel partners for implementation, methodology delivery, and managed operation of Lumen deployments. Through this network, enterprises get access to teams that have deployed runtime agent governance in regulated environments, from pilot to production.

What changes on day one

What has to change in existing agents and applications?

Nothing. No agent is re-coded. No application authorization is reworked. The Gateway governs traffic inline, the Broker front-ends generated MCP servers and A2A agents, and POET builds the agent surface from API specifications that already exist. Adoption is a deployment decision, not an engineering program.

Can enterprises start small and scale?

Yes. A common path: begin with discovery to see the agent estate and its risk scores, govern one PeerGroup, make one application agent-ready with POET, then expand. Helm and Terraform deployment makes each step repeatable, and the same control plane covers the first agent and the ten-thousandth.

Why not build this in-house?

Runtime authorization for agents means keeping pace with MCP and A2A protocol evolution, OWASP agentic threat research, SPIFFE and token exchange standards, and every agent platform's release cycle, indefinitely. Backflipt carries that burden as its product, with compliance to leading standards built in and no lock-in: generated MCP servers and A2A agent code export to hyperscaler AI environments such as Microsoft Foundry and Amazon Bedrock AgentCore.

What support does Backflipt provide?

Implementation guidance, training, ongoing technical support, and dedicated account management, backed by the partner network for specialized requirements. A twelve-year record with Fortune 500 production environments means support is staffed for enterprises whose transfers, agents, and audits cannot wait.

The Backflipt difference

No agent is re-coded.

No application authorization is reworked.

Enterprises scale AI on the identity and security investments they already have.

Topics

AI agent governance Runtime authorization Agent-to-agent security MCP security PeerGroups Agent microsegmentation Policy decision point Policy enforcement point SPIFFE mTLS RFC 8693 token exchange Non-human identity Workload IAM Machine identity Ephemeral credentials Credential broker Zero standing privileges Shadow AI OWASP agentic threats Agent-ready enterprise Trust No Agent MFT observability Air-gapped deployment

About Backflipt

Backflipt is the brand of Xenovus, Inc., an enterprise software company headquartered in Santa Clara, California. Backflipt builds the Lumen line of products for securing and governing AI agents in regulated enterprises: the Lumen A2A Gateway, an inline policy decision and enforcement point for agent-to-agent and MCP traffic; Lumen POET, a no-code platform for transforming APIs into governed MCP servers and A2A agents; the Lumen POET Credential Broker for ephemeral credentials; and Lumen AI TransferView for governed file transfer observability. Founder-funded since 2013, Backflipt has production deployments in regulated Fortune 500 environments.