By Enterprise AI agents are hitting a reliability ceiling that has nothing to do with model capability. The bottleneck is data: fragmented sources, missing integrations, absent semantic layers, and governance frameworks built for deterministic code, not autonomous agents operating across hundreds of systems simultaneously.
In this interview on TFiR, Mario Moscatiello, VP Growth at Airbyte, covers how Airbyte is extending its six-year foundation in data movement to support fully agentic workflows, including bidirectional data access, write-back to CRM systems, multi-interface agent tooling, and on-premise deployment for data sovereignty.
Guest: Mario Moscatiello, VP Growth at Airbyte
Show: TFiR
Here is what every platform engineer and enterprise AI architect needs to know.
Technical Deep Dive
Q: Is the AI model itself the main bottleneck for enterprise AI deployments, or is data connectivity the bigger problem?
Mario Moscatiello, VP Growth at Airbyte, argues that the model problem has been largely solved. New models ship every week, ranging from general to highly specialized, both open source and closed source. The real constraint is that even large enterprises often lack the data integrations their agents need, meaning critical business data is simply not available in warehouses or connected systems.
“The model problem has been largely solved. What you’re seeing is that even in the case of very large enterprises, a lot of the data that these agents would need to work on is actually not available in the data integrations or in their warehouses or in their systems.” — Mario Moscatiello, VP Growth, Airbyte
Q: How does missing data connectivity cause AI agents to hallucinate?
When agents lack access to a complete, consistent data foundation and a semantic layer, they generate answers based on incomplete context. Moscatiello notes that if ten people in a business ask an agent the same question, they may receive five or six different answers, a direct consequence of missing semantic alignment across data sources.
“Data connectivity is what’s causing a lot of agents to hallucinate because they think they’re giving the right answers to a problem, but there is no notion of semantic layer.” — Mario Moscatiello, VP Growth, Airbyte
Q: What is Airbyte building to address enterprise data connectivity for AI agents?
Airbyte is building a fully agentic platform on top of six years of data movement and integration work. The goal is to ingest data from many sources, organize it coherently, and push it back into the operational systems teams use every day. Moscatiello frames this as helping companies assemble context so their agents can reason and act reliably.
“We’re building a fully agentic platform where we are building on top of six years of data movement and data integration to say, how do we help companies assemble context so that their agents can work in a smart way.” — Mario Moscatiello, VP Growth, Airbyte
Q: What is reverse ETL and how does it relate to agentic AI write-back?
Reverse ETL, also called data activation, is the practice of transforming data in a warehouse and writing it back to operational systems such as Salesforce. Before AI agents, this was handled by deterministic DBT models running on a schedule, for example, identifying top spenders each morning and updating a CRM automatically. Moscatiello explains that agentic write-back extends this concept but replaces deterministic code with reasoning agents, which introduces new reliability and governance requirements.
“You could have a DBT model that runs every day and says who were my top spenders yesterday, and of course what you want to do is have a system that every morning writes that back into Salesforce so that the sales team can look at who were the biggest spenders.” — Mario Moscatiello, VP Growth, Airbyte
Q: What new governance challenges does agentic write-back introduce compared to deterministic data pipelines?
Deterministic pipelines are managed by a small data team, often 50 to 100 people even in a 10,000-person organization. When agents write back to systems on behalf of every employee, the governance surface expands by orders of magnitude. Moscatiello identifies three specific requirements: confirming the agent is not hallucinating before it writes, verifying that individual users are authorized to write to specific records, and maintaining full lineage and traceability so teams can reconstruct why an agent took a particular action.
“You go from a few people in an organization owning and governing access to data, to suddenly everybody being able to work on data. And so the scale at which that happens is just massive.” — Mario Moscatiello, VP Growth, Airbyte
Q: Why is data lineage and traceability especially critical in agentic systems?
In deterministic systems, errors are predictable and detection tooling is mature. Agentic flows are non-deterministic: an agent reasons its way to an action, and the path is not always visible. Moscatiello stresses that teams need to be able to trace back to why the agent made a decision, what data it acted on, who or what triggered it, and whether it ran on a schedule or autonomously. Without that traceability, debugging and auditing become impractical at scale.
“You need traceability. You need to understand, yes, this is a non-deterministic flow where an agent is reasoning and taking action, but you always need to be able to trace that back into why did the agent actually make that decision.” — Mario Moscatiello, VP Growth, Airbyte
Q: How does Airbyte support multiple agent interfaces such as MCP, SDK, and CLI, and why does that matter for enterprises?
Different roles in an organization naturally gravitate toward different interfaces. A marketing team member working inside Claude or OpenAI will use an MCP. A developer building a custom agent application will use the SDK. An engineer managing agentic workflows from the terminal will use the CLI. Moscatiello’s argument is that the interface layer can and should vary by use case, but the data infrastructure underneath must remain consistent so that a revenue question asked via MCP and the same question asked via SDK return identical answers.
“It doesn’t matter whether your teams are querying from our MCP or building something custom with the SDK or the CLI. We want to make sure that they’re reading and writing from the same data so that everybody in the business speaks the same language.” — Mario Moscatiello, VP Growth, Airbyte
Q: Is Airbyte’s MCP available on platforms beyond OpenAI’s marketplace?
Moscatiello confirms that Airbyte is working with all frontier labs to get its resources listed on their official marketplaces, with OpenAI having moved faster. In the meantime, the MCP can be installed directly into Claude by pasting a link. Marketplace listing provides discoverability and active support from the provider’s team, but functional access is not dependent on it.
“You can still go ahead and install an MCP into Claude. It’s just pasting a link. Of course, it’s nice to be in the marketplace so that people can find it quicker.” — Mario Moscatiello, VP Growth, Airbyte
Q: How much control do developers have over AI tooling choices versus top-down enterprise mandates?
In smaller companies, developers tend to have significant freedom to experiment across providers with usage-based pricing. As companies scale, cost governance takes over. Moscatiello cites the example of a large organization exhausting its annual token budget within the first few months of the year, forcing executive-level provider negotiations. Larger enterprises typically move toward top-down model selection for cost control, while still recognizing the need to remain model-agnostic to swap providers when a better model emerges.
“A lot of these companies are realizing that they also need to stay open to working with different providers and building all of their stack in a model-agnostic way, because if another provider comes up with a better model in three months, you should be able to swap everything and just use that model because it’s going to give you a competitive advantage.” — Mario Moscatiello, VP Growth, Airbyte
Q: What are the biggest technical barriers to running enterprise AI agents at scale in production?
Moscatiello identifies four barriers. First, integration plumbing: every data source requires solving for OAuth, pagination, rate limits, and error handling, and when agents scale to thousands running across hundreds of systems, a single API change can cascade into failures. Second, a reliability ceiling where most companies can only get agents working 80% of the time, which is insufficient for autonomous operations. Third, context degradation, where flooding agents with too much context causes them to lose signal and produce wrong answers. Fourth, immature security governance, where access controls built for deterministic code and small data teams are not designed for thousands of concurrent agents.
“The integration plumbing tax is real. Every data source that you’re trying to get into the model means that you have to solve for OAuth, pagination, rate limits, error handling, and all of that stuff.” — Mario Moscatiello, VP Growth, Airbyte
Q: Why is 80% agent reliability a hard ceiling and not an acceptable production baseline?
When a company relies on autonomous agents to run business operations, a 20% failure rate is not a manageable edge case; it compounds across every agent and every workflow. Moscatiello also flags that agent errors can be subtle enough to evade detection, unlike deterministic code where drift from expected outputs triggers established alerting systems. Closing the gap from 80% to 99% reliability is, in his view, one of the hardest open problems in production AI, and it is driving heavy investment in evaluation frameworks and regression testing.
“Working 80% of the time doesn’t really work at scale, especially because agents sometimes make errors that are so small that they’re not even detected.” — Mario Moscatiello, VP Growth, Airbyte
Q: What is context degradation in AI agents and how does it manifest?
Context degradation occurs when an agent is given access to too much data without clear prioritization. Rather than improving accuracy, an oversaturated context window causes the agent to lose the ability to distinguish relevant from irrelevant information, producing wrong or inconsistent responses. Moscatiello frames this as a consequence of the assumption that more data always helps, when in practice the semantic organization of that data matters as much as its availability.
“You think the agents should be able to see all of our data, but the agents can produce wrong responses because they have too much context and they don’t know what’s important and what’s not important.” — Mario Moscatiello, VP Growth, Airbyte
Q: How is Airbyte responding to enterprise demand for data sovereignty and on-premise AI deployments?
Airbyte has supported on-premise deployment for four years, a direct consequence of being open source. Moscatiello notes that while regulated industries such as finance, insurance, and healthcare have long operated under legal mandates that prevent data from leaving their environments, non-regulated software companies are now arriving at similar conclusions independently. Sending proprietary business context to frontier labs and effectively renting intelligence back is increasingly seen as a strategic risk, not just a compliance issue.
“Your first-party data is your gold, and as much as you can, you should always strive for data not to leave your environment.” — Mario Moscatiello, VP Growth, Airbyte
Q: How does Airbyte describe its strategic direction and what does the sovereign context layer mean in practice?
Moscatiello describes Airbyte’s trajectory as moving toward becoming the context layer for enterprise, with an emphasis on sovereignty. That means integration remains core, but the platform is expanding to help companies classify, organize, and make sense of their data so that all agents, regardless of the interface they use, read from a single consistent foundation that never leaves the organization’s control.
“We’re increasingly working towards becoming the sovereign context layer for enterprise where the integration is part of it, but there is a lot of work we can do to help companies make sense of their data, classify their data, organize their data, and make sure that everybody in the business is talking the same language.” — Mario Moscatiello, VP Growth, Airbyte
Q: What systems are next on Airbyte’s roadmap for agentic write-back after Salesforce?
Moscatiello identifies HubSpot as the next CRM target, noting it is already in progress. Beyond CRM, the roadmap includes ticketing and project management systems such as Zendesk, Linear, and Asana, where agents can write responses to support tickets or create and update project tasks autonomously. The longer-term goal is to make agentic write-back available across all of Airbyte’s 600-plus existing integrations.
“Airbyte has over 600 integrations on the data replication side and we hope that one day we’ll have all of those available for agentic products.” — Mario Moscatiello, VP Growth, Airbyte
Q: What is Airbyte’s view on open source AI models and model sovereignty?
Moscatiello expresses strong support for open source model development and flags concern that the US is currently behind China in producing competitive open source models. He predicts that model sovereignty will become a significant consideration for enterprises over the next five to ten years, parallel to the data sovereignty conversation, as companies look to avoid strategic dependence on any single frontier lab.
“I certainly hope that the US will start leading the charge with open source models because I think we’re lagging behind, especially compared to China. Companies are going to look more and more at sovereignty even with models.” — Mario Moscatiello, VP Growth, Airbyte
Resources & Documentation
- Airbyte, open source data integration platform with 600-plus connectors supporting ELT, reverse ETL, and agentic data access
- Airbyte Documentation, official documentation covering connectors, deployment options including on-premise, and platform configuration
- Airbyte on GitHub, open source repository for the core Airbyte platform
***
👇 Click to Read Full Raw Transcript
Swapnil Bhartiya: Every enterprise seem to have AI models now, and there are a lot of those. What they don’t seem to have is trusted access to their own data. That gap is where most agentic AI projects stall. They look very great on paper, but when you actually try to deploy it, that’s where things start to fail and fall. Airbyte is now tackling this by giving AI agents the ability to not just read enterprise data, but also write back into systems like Salesforce. And to unpack what this shift means for the future of enterprise AI, we have with us Mario Mascatiello, VP of Growth at Airbyte. Mario, it’s great to have you on the show.
Mario Moscatiello: Great to be here. Thank you.
Swapnil Bhartiya: Mario, first of all, it’s great to have you here on the show. And of course, AI or Genai or agentic AI, it has kind of built become the only topic that we talk here at tfi. No matter who I talk to, AI is the only topic. And what we have realized so far is that almost every enterprise, they have no shortage of AI models where the struggle is to actually their own enterprise data, enterprise context. If I ask you, based on your interaction, do you also feel that data connectivity is becoming a real bottleneck for agentic models as compared to the model themselves? What are you seeing when it comes to model versus data?
Mario Moscatiello: I think every week and every month we have new models that are coming out and they’re better and better. And there is. You go from general models to very specialized models, whether they’re closed source or open source. So I would say the model problem has been largely solved. I think what you’re seeing is that if you look at companies, whether they’re small companies or large companies and enterprise and organizations, they sort of have two things. You look at data maturity and you look at AI readiness. And yes, in a lot of the cases, companies that are highly data mature are usually the first to want to be embracing AI because they can trust their data. But in a lot of companies and at everybody we work from startups to Fortune 500, we see that even in the case of very large enterprises, a lot of the data that these agents would need to work on is actually not available in the data integrations or in their warehouses or in their systems. And so I think data connectivity is what’s causing a lot of agents also to hallucinate because they think that they’re giving the right answers to a problem. But there is no notion of semantic layer. If 10 people in the business were to ask an agent the same question, they would probably get five or six different answers. And so we’re really seeing that becoming a huge bottleneck for companies. And that’s definitely something that we’re here, you know, to help with.
Swapnil Bhartiya: So excellent. And what is Arbyte doing to address this problem?
Mario Moscatiello: So in a way where what we’re seeing is that because organizations have, you know, different levels of data maturity, what we’re building is really a system that can take a lot of data from different sources for companies and sort of like organize it in a way that makes sense. And then helping them, companies helping those companies put that bit, put that data like into action and put that data where it needs to work so back into the systems that they use every day. So we’re building a fully agentic platform in that sense where we are building on top of six years of data movement and data integration to say, hey, how do we help companies assemble context so that their agents can work in a smart way.
Swapnil Bhartiya: Can you talk about how it just kind of pushes your byte beyond just retrieving data into actually now you are not just looking at it as not just one way traffic, you are actually also taking actions like writing back to very sophisticated system like Salesforce. How does that bidirectional access change the way enterprise agents are architected and what new governance challenges come with it? Because if you do allow your agents to start taking actions to start writing to your surface system, you need to have right governance as well. So it is not as that simple. So can you talk about the working of the cli?
Mario Moscatiello: I think the CLI is one component but where we allow developers to just interact with our platform directly from the terminal. But I think what we’re talking about here is more like us enabling agent to write back to systems. If we look back at the last five years, this has already happened when it comes to just pure data. So I’ll back up a bit. But essentially we also have this feature, but there are a lot of companies that a bunch of companies that have developed this feature before, which is called what we call reverse CTL or data activation. And what that means is that this is before AI. You could have a DBT model that runs every day and says, hey, who were my top spenders yesterday? And of course what you want to do is have a system that every morning writes that back into Salesforce so that the sales team can look at who were the biggest spenders and try to upsell them or trying to lock them into a contract as an example. And so this reverse ETL actually transforming data and writing it back to systems is not something that is just out today, it’s something that companies have been doing for a while. I think the issue is that when you go from a data team writing very deterministic code that is going to run every day and just update, you know, stuff in a CRM or in whatever system, when you go from that to an agent writing back to a system, first of all you need to make sure that the agent is not hallucinating and then the agent is writing the right data to the right system. And second, you go from a few people within organization, even if it’s a large organization, you can team of, even A company of 10,000 people is going to have maybe 50 people, 100 people in the data team if it’s already very data mature. When you go from 50 people to suddenly everybody’s agent having access to writing back to systems, how do you know that a specific person is even allowed to write back to their record? Or how do you know that a specific person has access to the information he needs for the agent to take action? And so from a governance endpoint, it requires way more robust governance and so on and so forth. But you also need way better than the lineage because you need to be able to understand like why did an agent take an action? You need traceability, you need to understand, yes, this is a non deterministic flow where an agent is reasoning and taking action, but you always need to be able to trace that back into, okay, but why did the agent actually make that decision? Why did the agent wrote that data? Who initiated that agent? Was it on a schedule, was it automatic? And so on and so forth. So I think you’re going from a few people in an organization owning and and governing access to data, to suddenly everybody being able to work on data. And so the scale at which that happens is just massive. And so it introduces a lot of challenges.
Swapnil Bhartiya: It’s available on OpenAI Marketplace. What about other like cloud and Tropic cloud? Is there other? Yeah, we’re working with a lot of cloud.
Mario Moscatiello: Yeah, we’re working with them. You know, it’s a process. Like every company. OpenAI was quicker in this case, but it’s a process. And we’re working with all of the frontier labs to have these resources officially on their marketplaces. You can still go ahead and install an MCP into Claude. It’s just pasting a link. Of course, it’s nice to be in the marketplace so that people can find it quicker. And their team is also doing some work on it, but we do work with all the modern providers.
Swapnil Bhartiya: Of course, if you look at AirByte and I have been covering AirByte from the very early on, so I have seen your journey. Now you folks are giving agents multiple ways to interact with Enterprise Systems, NCP, SDK, now CLI and of course availability in OpenAI App Marketplace. And it will be available in other places as well. How do you see these interfaces evolving? The AI market is changing very fast, so it’s hard to say that, but just, you know, based on your interaction, based on your own usage. Because today everybody is using AI either way, do you see one will win out as the standard or there will be a fragmentation, there will be a mess where enterprises will have to deal with all of them depending on the kind of agents they need. Also, a lot of players, even anthropic, they have launched managed agents because they all understand that agents are becoming another big part of the problem for enterprises to manage.
Mario Moscatiello: I think you hit the nail in the head, right? I think it’s different people within an organization will use different interfaces for different types of work. I think if you’re in a marketing team and you’re using Claude or OpenAI and you want to use an MCP because that’s the interface, that’s your interface for work, that’s great. If you’re a developer and you’re building an agent and you need the data SDK and you’re using the SDK, that’s great. You’re building a custom agent for a specific thing or you’re building an application that has agents in the back, that’s great. And if you’re again an engineer working out of your terminal and you want to have access to a command line interface to manage everything you do in this agentic application, that’s okay too. And I think that what’s the interesting thing is that something like we do with airbyte agents is we can say, look, we know that enterprises need access to different interfaces for different classes of agents, but what shouldn’t change is the data infrastructure behind them. So that if your marketing team is using the MCP and they’re querying data, or your developers are using the SDK and the cli, they should read from the same data. Like if you know at the end of the day, if the marketing team is marketing person is building an agent, that should answer how much revenue did we do yesterday? And an engineer doing it with the MCP and an engineer is trying to do the same with the SDK because they’re Building a revenue agent. To that question, the agent, whether it’s a question, is asked from an NCP or from an SDK, the answer should be the same. And so what we can say is, look, it doesn’t matter whether your teams are querying from our MCP or building something custom with the SDK or the cli. We want to make sure that they’re reading and writing from the same data so that everybody in the business speaks the same language.
Swapnil Bhartiya: How much say developers have when it comes to AI versus how much is the top down? Because that also dictates the kind of framework you are building, the kind of tools you’re using, the kind of AI you’re allowed to use or not use.
Mario Moscatiello: I think in smaller companies developers have the freedom to experiment. Usually know smaller companies are more flexible and they tend to have, you know, more usage based pricing when it comes to like, yep, you can, you can experiment with a bunch of different providers and so on and so forth. I think when, when the companies like start scaling, you have, you know, companies like Uber saying wait on a second, like it’s, I don’t remember the exact month, but they were saying, hey, it’s March or April and we’ve already used all of our token budget for a year. We need to put some governance in. And so in that sense, if you have the board and the executives, they probably go and negotiate with one of the providers and say, hey, if we were to deploy your models across the entire companies, what is the pricing? And so I think from a cost perspective, larger companies will tend to have more top down in that sense because these models are expensive and these providers are expensive and so they need to put it in place. Again, we go back to governance. It’s not only governance for the data side, but it’s also governance for how much they’re spending and how many tokens they’re using and are they getting roi. But I think what’s happening right now is that a lot of these companies are realizing that they also need to stay open to be working with different providers and building all of their stock in what we call like a model agnostic way. Because you know, for a company, if in three months, if you’re working with one provider and another provider comes up with a better model in three months then is way better for you. You should be able to in theory swap everything you’re doing and just use that model because it’s going to give you a competitive advantage. And so I think it’s kind of like a fine balance in between cost and sort of like flexibility. And that’s what we’re also seeing a lot of the open source models really, really taking shape. And you have some great open source models that are being developed. I certainly hope that the US will start leading the charge with open source models because I think that we’re lagging behind, especially compared to China. They have some amazing labs, they’re coming up with amazing models. And I think that companies are going to look at more and more, are going to look at sovereignty even with models. So it’s going to be an interesting five, ten years ahead.
Swapnil Bhartiya: As we all know that a lot of companies are experimenting with AI agents, but going from a slick demo that they saw in the boardroom or with developers to production system is still hard. Can you talk about what are some of the biggest technical hurdles that keep enterprises from running thousands of autonomous agent at scale?
Mario Moscatiello: Yeah, I would say, like, of course I would say in the case of Airbyte, what we see is that the integration plumbing tax is real. Every data source that you’re trying to get into the model means that you have to solve for oauth pagination rate limits, error handling and all of that stuff. And so it’s stuff like the more agents you have running, the more systems they’re reading and writing to. So if you now have agents writing back into Salesforce, the moment Salesforce updates their API, it might break something. And so you have to. Integrations are still know like the number one issue in, in that sense in, in based on what we see because it’s very hard. And if you have a thousand thousand agents that are reading from hundreds of systems, it’s a multiplier effect. And so like that’s the number one cause that we see today. I would say the second thing is that at most companies are able to have agents work 70, 80% of the time, which is kind of like a bad ceiling. Because when you’re running, if your company’s been like running autonomously or if you want your company to run autonomously, you know, working 80% of the time doesn’t really work at scale, especially because I think that agents sometimes make errors that are so small that are not even detected. And so, you know, engineering has built a lot of guardrails when it comes to errors and when it comes to responses that are not what they expected. Especially with deterministic code, right, you run code and you expect a response if there is a drift. Like there is systems that are put in place to detect even the smaller errors when it comes to agents, it’s a completely different pair of ants. And so I think that 20, 30% of like hey, can we get the agent to work 80% of the time to working or 90% of the time to working like 99% of the time. I think that’s going to be a hard problem to solve. And that’s kind of like why a lot of you hear a lot about evals, making sure that we can look at regression and the performance and so on and so forth. So I think that’s the second big problem. I think the third one is context degradation which is when you just flood the agents with a lot of context and you think like oh, the agents should be able to see all of our data and so on and so forth. The agents can produce wrong responses because they have too much context and they don’t know what’s important and what’s. And what’s not important. And so I think to sum it up, those are the three main causes. But the fourth one to your point earlier is governance. And we go back to security. Access has been built for deterministic code and for a few people within an organization managing highly secure information. But Security Guard lets governance is not built for thousands of agents running. And so I think the security stack is immature. And so that’s going to be also like a big, in my view, a big point to solve.
Swapnil Bhartiya: Now there’s a lot of geopolitical crisis going on and decisions are being made which may or may not be good for AI. A lot of anti AI fuddies also going on. So we are living in an era where there are a lot of advantages and then also there is a lot of skepticism is also going on there. Regardless what we are seeing is that more and more organizations, more and more companies, more, more and more countries, they do want AI sovereignty. How does Airbyte facilitate that? Because we are also, you folks also work in regulated industries, compliance sending industries and also when you throw the whole of course in eu a lot of AI related laws are coming into force as well. So talk about how is Airbyte helping organizations countries to take control of their AI inside out.
Mario Moscatiello: Yeah, I think something we’ve been saying a lot for the past two years at Airbyte especially is that when you’re working in these sensitive applications, but also when you’re not like now, it’s like anybody, you should, you know, your first party data, it’s your gold essentially as a company and as much as you can, you should always strive for data, not for data. Not to leave your environment essentially. And now this has been true for regulated industries simply because they didn’t have a choice. If you’re a finance organization or insurance or healthcare, well you have no choice. Your data by law cannot leave your environment. But when you’re a software company like Cloud Native, maybe that matters a bit less. Until now because what we’re seeing is that a lot of companies are saying, yes, we’re giving Frontier Labs and all of these providers all over our context and then we’re kind of like renting back our intelligence in that sense. And so like we, you know, we have hybrid deployments, we’ve been deploying on plan on prem for four years because we are open source. And so what we’re seeing is a lot of companies even that are from non regulated industries are coming to us and saying hey, like do you have a system for which I can give data to my agents without that data leaving my environment? And that’s sort of what we’re seeing today and that’s what we want to help with and that’s what we want to like, you know, focus on is how do we provide governed data access that is fully within an organization control so that, you know, companies are not really selling their secret sauce.
Swapnil Bhartiya: I would not ask you how would Airbyte look like five to six years from now? Because we don’t know what kind of technology will come in.
Mario Moscatiello: We don’t know in six months.
Swapnil Bhartiya: Exactly. But if I ask you that since the whole inception and arrival of Airbyte and if you look at today, how would you explain what is Airbyte today to organizations? What kind of company are we looking at?
Mario Moscatiello: Yeah, I think we’re increasingly working towards becoming the context layer, you know, for, for enterprise. And I think that that’s definitely like one aspiration that we have. And I would say like based on, you know, the last point that we were discussing right now is can we become the sovereign context layer for, for enterprise where the integration is part of it. But there is a lot of work that we are doing or that we can do to help companies make sense of their data, classify their data, organize their data and again like making sure that the business is, everybody in the business is talking the same language and that these agents can extract very useful information from the data.
Swapnil Bhartiya: Now with Riot Access now live for Salesforce, what other systems are next on the roadmap for agents to take action on? Not just to read from, just tease us, don’t share too much, we can talk about it when it’s ready. But just tell us what next HubSpot
Mario Moscatiello: is definitely like, you know, we already have it working. It’s kind of like next on the list. I think CRM is the biggest place where a lot of business users write their data. And so I think that’s a very interesting one. And I think next you can think of product type systems or ticketing system, whether it is that is Zendesk or Linear or Asana and those type of tools where there’s a lot of project management happening and you want an agent to write a ticket or an agent to help you solve a ticket. And so if you’re building a support system, for instance, like you want an agent to be able to write back a response to Zendesk to a customer or something. And so those are a few of the systems that we’re looking at. But Airbyte, we have over 600 integrations on the data replication side and we hope that one day we’ll have all of those available for Agentic products.
Swapnil Bhartiya: So, Mario, thank you so much for joining me and walking us through where AirByte is headed with agents that don’t just read data, but actually act on it. And those who are watching, please head over to airbyte.com to check more about what they are doing in the space. Thanks for watching. I’ll see you next one. Thank you.
Mario Moscatiello: Thank you.
