By AI agents are useless without access to enterprise data, but unrestricted access breaks compliance, violates data sovereignty requirements, and makes audit trails impossible. Write-enabled agents acting inside production systems with no permissioning layer create accountability gaps that most enterprises cannot accept. These three problems block nearly every serious enterprise AI agent deployment before it reaches production.
In this interview on TFiR, Jean Lafleur, Co-Founder and Chief Operating Officer at Airbyte, breaks down exactly why enterprise AI agents stall on data access, sovereignty, and permissioning, and walks through how Airbyte’s hybrid deployment architecture addresses all three.
Guest: Jean Lafleur, Co-Founder and Chief Operating Officer at Airbyte
Show: TFiR
Here is what every platform engineer and AI infrastructure team needs to know.
Technical Deep Dive
Q: Why do AI agents fail to reach production in enterprise environments?
Jean Lafleur, Co-founder and Chief Operating Officer at Airbyte, identifies three structural blockers that prevent enterprise AI agents from moving beyond demos. First, data sovereignty rules mean enterprise data must remain inside the organization’s own environment and cannot travel to external services. Second, agents that lack access to internal data are too limited to complete real tasks. Third, write-enabled agents operating without a permissioning layer create accountability gaps: there is no record of who authorized an action, on whose behalf it was taken, or which agent made a change to a production system.
“Agents are very limited if they don’t have access to the data.”
Jean Lafleur, Co-founder and Chief Operating Officer, Airbyte
Q: What is the data sovereignty problem for enterprise AI pipelines?
Many enterprise environments operate under regulatory or internal policy constraints that prohibit sensitive data from leaving their infrastructure. When AI pipelines route data through external cloud services, including third-party LLM APIs or SaaS connectors, those policies are violated. Lafleur notes that unless an organization runs fully local models, every API call creates a potential sovereignty exposure, making cloud-only data pipeline architectures incompatible with enterprise compliance requirements.
“A lot of the tools, the data is within the enterprise environment and they want to have the data stay there.”
Jean Lafleur, Co-founder and Chief Operating Officer, Airbyte
Q: How should enterprises handle permissioning when AI agents have write access to production systems?
Write-enabled agents present a specific governance challenge: any change made to a field, record, or system must be attributed to a human principal, not just the agent process. Lafleur frames this as requiring two layers of permissioning. The first layer controls which data an agent is allowed to read, scoped to what the delegating employee is authorized to access. The second layer creates an audit trail recording which agent made a change and on whose behalf, so that accountability is preserved even when the action was automated.
“If an agent changes a field somewhere in a tool, we need to know who has done that on whose behalf.”
Jean Lafleur, Co-founder and Chief Operating Officer, Airbyte
Q: How does Airbyte Flex solve data sovereignty without sacrificing cloud support?
Airbyte addresses the sovereignty problem through a hybrid deployment model called Flex. The control plane is hosted by Airbyte and handles orchestration, support, and pipeline management. The data plane runs entirely inside the customer’s own environment, meaning Airbyte never has access to the underlying data, only the metadata required to operate and support the pipelines. Lafleur describes this separation as the key unlock for enterprise adoption, because it satisfies internal data residency requirements while still allowing Airbyte to deliver a managed service experience.
“The data plane stays within your environment. We have never access to your data, only some of the metadata that helps us control and provide you support for your data pipelines.”
Jean Lafleur, Co-founder and Chief Operating Officer, Airbyte
Q: Does the Airbyte Flex hybrid model extend to agentic AI workflows, not just data pipelines?
Lafleur confirms that the same Flex architecture that governs data pipelines applies directly to agent workflows. The data plane remains in the customer environment, which means agents operating through Airbyte never route sensitive data outside the organization’s perimeter. This makes the same hybrid deployment model that enterprises already use for pipeline compliance the foundation for governed, production-ready agent deployments.
“Same for agents, agent workflows. That’s how Airbyte solves the problem by having that hybrid solution.”
Jean Lafleur, Co-founder and Chief Operating Officer, Airbyte
Resources & Documentation
- Airbyte, open-source data integration platform with hybrid deployment via Airbyte Flex for enterprise data sovereignty and governed AI agent workflows
***
👇 Click to Read Full Raw Transcript
Swapnil Bhartiya: Let’s start with that problem area. When it comes to enterprises, AI in production, what kind of friction? You see, at one hand, agents do need access to knowledge, otherwise they won’t be able to perform tasks at the same time. Now they also have write access. They can do whatever they want. So you also need to govern what they can or cannot do. So let’s talk about the problem area. Then we’ll talk about what Airbyte is doing.
Jean Lafleur: At the enterprise level we see three main problems. One, the first one is sovereignty, like data sovereignty, a lot of the tools, the data is within the enterprise environment and they want to have the data stay there. So that’s one problem. The second one is yes, agents are very limited if they don’t have access to the data. And as you mentioned, the third one is really permissioning. And permissioning meaning, okay, who has access to this data? Because while an employee or an agent from that employee shouldn’t have access to some data and also the right permissions so that if an agent changes a field somewhere in a tool, we need to know who has done that on whose behalf. And those three problems are pretty complicated. And Airbyte Agents, that’s what we are trying to solve.
Swapnil Bhartiya: I focus on two problems. One was access to data and second is what they can do with the data. And you also talk about sovereignty, which is actually becoming very important because if you unless you’re using everything local LLMs, then it’s not a problem. Then you have to just focus on which box. But if you’re using API, if you’re a front end model or open source model, then you do lose. So talk about that aspect. Also when we talk about Airbyte and of course your Airbyte Agents platform, how do you tackle these problems? And then we can talk about how the latest update is actually making things better.
Jean Lafleur: What we have is a hybrid deployment solution that we name Flex. And that means the control plane is hosted by Airbyte, but the data plane stays within your environment. So we have never access to your data, only some of the metadata that helps us control and provide you support for your data pipelines. And that’s the way we seen no enterprises having any issue with this. That’s a big unlock for them. And same for agents agent workflows. So that’s how Airbyte solves the problem by having that hybrid solution, we provide the best, meaning you have the data plan saved in your environment and also we can provide the best experience because we can provide you support.
