Name: Spring Platform Highlights
Uploaded: 2026-06-02T19:16:13.209Z
Duration: 59 min 1 s
Description: Spring Platform Highlights

Transcript for "Spring Platform Highlights": Alright. I'm gonna go ahead and get us started. Welcome, one and all, to the spring Celigo platform highlights. Many of you have joined sessions like this in the past. We changed the format a little bit, and we're going to look back in the recent past to let you know some of the things we've released and have a little look forward. I've been joined today by Nate Briant from our product marketing team and Tyler Lamparter from our product team. I'm Matt Graney, chief product officer at Celigo. Many of you, I've I've had a chance to address in previous, sessions like this. I'm really glad to have you with us. We've got a packed agenda today. But before we get into that, I just wanted to set the stage. You know, we are gonna be talking a bit today about AI. No conversation seems to be complete now without a mention of AI. And I'd encourage you all to to check out this MIT report we, commissioned. It's available from our website. When you see the slides, the link will be available there as well. And what that study showed across a a wide spectrum of companies is that there's a strong correlation between those having success with AI in production and those that have an integration platform. And I don't think that's surprising because we know that AI needs data and AI needs APIs, data to understand, APIs to take action, and that, of course, is right at the heart of what an integration platform provides. So here at Celigo, we've been building on our integration platform for a long time and finding new ways to automate how businesses work. Many of you will know that we began very much in the A to A, the application to application space, and have gradually expanded. So we first added b to b, now formalizing our b to b manager product. We've always been an API first platform, so adding some API management capabilities was inevitable. And then, increasingly, AI has featured in our platform, both in, you know, how we manage errors in the platform and now how we provide AI in the user experience and now allowing customers to support AI workflows. We're also going to be mentioning something exciting coming in the data ingestion space as well. The net of this is one organic platform. We have built this all in a seamless organic fashion. The platform has very elegantly, I would say, and organically expanded to support this wide range of use cases all in one place. Because there's this single user experience, it gives us a lot of additional opportunities to streamline the way that work gets done in the platform to provide that sort of operational visibility that you need, to be able to deploy AI and, indeed, you know, automate other parts of your business at scale. So we've got a packed agenda. When we get into it today, you'll notice there are some things that are gonna be forward looking. So our lawyers, of course, always want us to put the right kind of disclaimer. Our plans are all well intended and the best known plans today, but they could change. So please do keep that in mind. And aside from that, I'll get out of the way and hand it over to, to Nate Briant from product marketing, and, we'll get us started. Thank you. Over to you, Nate. Thanks, Matt. So, we're gonna start by talking about the exciting ways Celigo is enabling users to add AI directly into their their operational workflow. So how do you use it today, to make it have real impact? But first, we wanna start with a poll. Right? If you could just respond on a quick survey about how how you are currently using AI. What's your organization's current use? How maybe you're using it internally? And the reason we're asking this is we know everyone is really on a different place in their AI development. So it kinda refer back to that circle that Matt talked about where the the goal of any platform to be a platform and really Celigo and our platform is to be able to support any of the use cases that our customers have no matter what form factor it might be take might take. So, and we wanna ensure that no matter the form factor, we're staying consistent with the strategy of our platform, which is making it accessible, reliable, easy to use for the automations that run your business. And in the past, right, that Matt talked about, that was mainly, integrations or task automations. But now with LLM agents, this has really expanded the tint of what automation use cases can be, and we wanna support that wherever wherever our customers need. And so the first aspect of that is is agent builder. So, this allows, our customers to configure agents directly in the interface. It uses simple natural language instructions. And a a big part of that is agents need access to the LLMs for the reasoning that they're gonna be doing. So that means interpreting the instructions, making decisions, and we wanna make accessing those models as easy as possible. So you can bring your own model with, if you already have a relationship with your model provider, or you can really get started quickly with our embedded models, which gives you access to those premier frontier models without needing a separate contract. You're already contracted with Celigo. You're using us for the integration automation. It makes it easy to get started and start using those embedded models with with the tokens you have. And, of course, the all of these agents, they're built directly on the Celigo platform where, you already have access to the secure connections. Your data is already moving through it. It has the access and the governance you need. So it's a natural place for these agents to to occur. And so what does an agent look like on on the platform? Right? So this is a a simplified diagram of something that we're all familiar with, which is an integration flow, and and you're can build a flow around an agent or invoke it in an existing flow. And and the agent contains just a few components that are easy to configure. So you provided the instructions and natural language. You give it the the model that, interprets those instructions and makes the decisions. The model you choose is is either your own that you're providing with your key or the embedded one. That's really best situated for for the task. Right? So it's very flexible, and you can, change the model as needed based off of the task. And then if the agent is taking actions, you need to give it the tools. So you, attach these tools, whether it's a Celigo tool or an MCP server that, has the tools underlying it so that the agent can take real action, whether it's on the data or the systems that you that you you want it to interact with. And then very critical kind of wrapping this is making sure that you have the guardrails for it. Right? The reasoning is being done by these models now instead of instead of a human. So how do you make sure that the correct data is exposed? One, that's through the access you've given through the tools, but also with the guardrail so you can filter out data that, shouldn't be distributed or moved by the agent. And so, this is an example of how we're using agents internally at Celigo with our own product. It's all about making sure that the use case really lessens the manual burden on on a team. So particularly in finance, they're receiving lots of emails to a a a shared inbox, and an agent can read these unstructured data coming in as an email, understand the request, and then take action based off of that. So it can common examples are requesting an invoice be resent or updating a a billing address. And so the agent's able to take that, access the email, and then either, create an email with an invoice attached or, recommend an update change within the database. And really critical amongst all this is that, especially in a finance use case, you wanna make sure that the agent is is operating correctly, but, also, you have a human in the loop. Right? So in this case, our agent is actually just drafting an email. Someone in finance can go and review it, make sure that it it's what what they need, and either make a change, delete the email, or send it. And and, really, what the agent is doing is it's it's doing the very tedious heavy lifting that that task required. Right? Reading an email, understanding the context, looking up the data. And so that frees up our our finance team to work on more strategic projects. They're not spending endless times searching for data and just responding, and it also gives a better customer experience. So it's all about creating, an agent that has real operational value and and is done in a controlled way. And that and only through the control plane of the of our platform with the right access, with the right controls is that able to happen. And so really critical for all of these agents is the idea that they have the access. So I'm gonna pass it over to Tyler to talk about MCP or model contact protocol, which is how these agents are accessing this data and performing the actions. Yep. Looks nice. Yeah. So no agent is no agent is, good if you don't have tools and assets available to that agent when it needs to run. So that's where MCP servers come in. So MCP servers allow agents to access, you know, various tools, resources, prompts, etcetera, so that agents can perform actions on your behalf. So agents themselves can have context to do things for you, and then it's got sets of tools where it can read data, write data, etcetera. And so on Celigo, given that we're an iPaaS platform and we've historically already been connecting with all of your hundreds or thousands of applications, right, you can expose all those applications to LLM clients like OpenAI, Gemini, Claude. Right? And from that, they can perform actions on your behalf. So you have a central governance layer within Celigo where you'd manage all those MCP servers. They're accessing your, data that's in Celigo, by calling, all of your various applications that you've already are connecting to within Celigo, like your ERP systems, your CRM systems, your support ticket systems, your sales, Gong, Zoom transcripts, and and whatnot. And what Celigo has done over the past quarter is we've added, tool publishing. So we've added a new resource type that we'll go through a little bit later in the call in more detail, but tools that you can attach to MCP servers where you've got, context for its input and context for its output, and the tool does exactly as you described. It's not gonna write data to some system or read data from some system. Additionally, because Celigo is gonna be the governance layer here for MCP servers, we've added in a new Celigo identity for MCP servers. Whenever you go to authenticate to an MCP server within Cloud or Gemini, you're directed to Celigo to authenticate against your Celigo user persona or your your user profile, and then you come back into Cloud and you were an authenticated user because you have that, access granted to you within the Celigo environment. And then lastly, we've implemented MCP prompts so that you can have reusable prompt assets whenever you're using workflows, like sending emails and you need, like, formal responses written the same way each time, or you need to know you need to give LLMs context into how to use the sets of tools that you've made available for them. And with that, Nate's gonna go through and see if there's a little bit more. Yeah. Awesome. I mean, that that's exactly why you would would you use MCP service for it. And so so why would you use Celigo in particular for MCP? We know that lots of SaaS companies are offering MCP specific for their application, and allows you to connect to that data. So it really comes down to the connectivity and control that Tyler mentioned because Celigo, lets users build their own tools that are then serviced through the MCP. You can create tools that otherwise would not exist. So in the case where apps don't have MCP servers, you can build those tools to have that functionality. But in the case where MCP servers do exist, there's a concept of, like, a composite, tool where it's not just one tool that you're interacting, one application. You may need data and insight from multiple tool from multiple sources to be available in that tool. So, you can think about designing, an MCP server around a particular persona. So if it's someone that's in marketing that needs access to Salesforce data, Marketo data, HubSpot data, you wanna offer all of that data up into a an MCP server that is is available for them to do their job. Right? This is especially powerful when you're trying to generate insights or or work with these applications, especially, when the in your single client. You can have all those tools available and giving the right access to the right people. So making sure that, if a tool is is not designed for that persona or you don't want them to have access to that data, Celigo, again, gives you that ability to control how that those tools work. And you can build them very quickly. Right? The faster time to value, you can be you can set it up on on the platform. You have that governance. And it's really future proofing this where, based off of the personas that need these the access to this data, they'll have the tools available to them. And so now we're gonna shift gears a little bit. So we're talking we've talked a lot about, like, AI products in our platform, but we're gonna talk more about how AI is working with builders. So how do you uplift the experience on the platform, using the AI tools that we're all, you know, becoming very prevalent, in in today's, today's world. So in in kind of the age of of of vibe coding, working with LLMs and agents, you can quickly prototype and build. It's becoming very common. I think people are building entire apps themselves. But on our platform, we want users to do more than just five code. Right? It's building real tangible assets, iterating on them, building in a secure way, and then has actual impact. And so to do that, every interaction that we build, that is done on the platform is done through our our Celigo APIs. And so this gives builders the functionality they need not only to build but manage. And and oftentimes, users don't even really know that's happening behind the scenes. So that's fine. We wanna meet a builder where they are. And so the first way we did that is by offering Celigo Aura. So that's augmenting the traditional UI experience where you're working in the platform, and now you have this extremely powerful agent that is able to assist you in what you're doing. So it can help you build entire flows. It can help you build subsets of flows where maybe you're struggling with a particular part. You can say, I need help with this transformation or help me with this mapping, and it's just speeding up that that build experience. And then it's going beyond that. It's helping you manage the your your instance. You have the ability to troubleshoot errors. You have access to the entire knowledge base. You can ask questions of, and it really is is, again, augmenting the UI experience that everybody's familiar with. And so it's an extremely powerful tool, and I'm gonna pass it over to Tyler to talk a little bit more about, about Aura and some of the use cases that that we've seen, really help our customers. Yep. Thanks, Nate. Yeah. So if you guys haven't checked it out yet or seen it, in the bottom right hand corner of your Celigo UI is a new button for Celigo Aura. And what this does is it brings up a conversational UI, over on the right side, and you can kinda drag it around wherever you need. But it's Celigo Aura where you can interact with almost any Celigo resource. So, we've been building this out over the past many months, and you can do tons of stuff with it. You can build flows with it, help you, troubleshoot errors. You can help it write mappings or handlebar expressions for you that, you know, historically might have been a little bit difficult to do. Maybe you need to do some complex transformations or scripting along the way. It can do basically all of that for you. All you have to do is have a quick conversation with it. It knows where you're at in the UI, so it has that contextual awareness of exactly where you're at to perform that particular action on your behalf. But even outside of that, it can, really help with, like, managing and monitoring errors. It's It's like a really good use case for it. So, you know, if you've got, you know, tens, hundreds of flows going on throughout the platform, you know, they're going to have issues from time to time. You may you know, maybe an order is missing a particular field that needs to map into, your ERP system or, you know, maybe, a a time out request or something, right, for a particular endpoint that you need to automatically retry. So with Aura, you can just simply go to Aura chat, and you can say, you know, what areas do I have in my account? What are most prevalent? And then it can automatically retry for the retry those errors for you. It can summarize those errors for you to see if there's any actions that you need to take. A really cool example I ran into the other day was, I had a situation where my script didn't handle, the incoming data exactly correctly. And so, you know, this flow would have been running for a month or so. And And then I had a new scenario crop up from my source application, and so the script failed because it didn't handle, it correctly. What Aura did at that point was it analyzed the error, read through the script, fixed the script, and then retried the error all within a single chat. So, I didn't have to go copy that script into a separate system and then troubleshoot there, then go click through the UI. So it's just a really powerful tool to go through the UI and manage anything and everything within Celigo. Awesome. And with that, I'll pass it back to Nate with some other cool opportunities we've got for, interacting with Solio. yeah. Totally, Tyler. And and and and it's really exciting to hear, like, an error that maybe would have taken time to research and and figure out and change, is really done all within the UI, through Aura. But it's also maybe moving in the future, right, in this world where everybody's wanting to kind of centralize on a central tool, whether it's an LLM client or LLM chat. It's making sure that we have the ability for those users to work in in in that way as well. So the next iteration, right, is is CLI. So this is for users that are really comfortable, in terminals or code editors where, these code editors and terminals themselves are really being augmented by AI. So with CLI, you have or command line interface. You have a library of commands. It's added as a package to your editor, and those commands are directly, again, tied to the APIs that allow you to take actions. And so with natural language, you're able to, put in a request within your editor, describe what you wanna do. The AI interprets that request, and then it identifies from the the commands what it would need to do to accomplish that task. And so whether it's a single command or a set of or a series of commands, it does its best to achieve what you want to happen. In a very similar vein, there's Celigo MCP, Platform MCP. And so this connects, to your LLM client and exposes a series of tools rather than commands. But, again, it's using LLM to interpret your request, what you're trying to do, and then, choosing the right tool based off of the API to accomplish what you need. And so both CLI and MCAP are part of a broader conversation around what what's being deemed headless operations. The idea that you're working inside of your AI enabled app, your client, where you do most of your work, and you're not having to switch between UIs. And so while it's headless and you're not working in UI, this is very different from from Vibe coding on a on a solution on the fly because the operations, the assets, like mentioned before, they're created on the application that you're used to working within the UI. So it has the performance, the security, the government governance, and you're just working in an interface that that you prefer. And so in the world of headless operations, CLI, MCP, they they they seem like very related approaches on the surface, very similar outcomes. But, and we wanna allow users to choose which one they prefer. But Tyler's gonna talk a little bit about the nuances between the two, why you might use one, some of the pros and cons around each. Yeah. Thanks, Nate. Yeah. So we've been iterating and moving very quickly over the past couple months for, bringing CLI and MCP PlatformMCP to all users. But first off, we released, CLI about maybe a month ago or so. And with CLI, really, it's meant for kind of we see them in two different personas. Right? So CLI really meant for, like, real, you know, developer heavy. They're in quad code. They're in codex. Things that would run bash commands against Zwigo. So list out connections, list out flows. And, additionally, maybe, like, partners who are accessing multiple accounts in multiple environments, you can have different profiles set up between, within the CLI, and then just reference that profile when you're running commands as opposed to, like, a platform MCP where typically you would have to set up, one MCP server, like, per account and per environment. So it's a little bit the two are a little bit different, but two completely different personas that are interacting with them. The CLI would be a little bit more efficient on token usage when you're running commands compared to, MCP. But MCP itself is more prevalent throughout the market. So, you know, the vast majority of LLM clients, right, can connect to MCP servers. They can't run CLI commands or bash commands if you're not local. Right? Maybe you're in a web browser and you need to connect to Celigo. You wanna be running CLI at that point. You would be running MCP server, tool calls. And so with that, the MCP server the platform MCP server that we're releasing here in probably the next month or so, will allow you to really manage errors within Celigo, fetch flows, build flows, all the stuff that you could do within the CLI, but all within the MCP server context. So we're just letting users, you know, build wherever they are, and giving all the tools available and all the resources that they need to build and accelerate on the platform. And and, Tyler, when we talk about these two options, obviously, these are extremely powerful. Is there you know, without being negative, is there is there dangers to using CLI where you could using natural language, it could misinterpret your request. How do we wanna mitigate that for for customers to make sure that, they don't end up maybe deleting a flow or something like that? Is that where you would use MCP versus CLI or or you can start using personal access tokens to make sure that you they have the right access? Yep. That's a good question. Right? So, currently, to authenticate against the CLI and the MCP server, you need you almost need, like, a full access token within Celigo. So you'd have to be an admin with a API token that has full permission sets. So it's, you know, Matt Graney's analogy analogy is typically, you know, it's like, giving somebody scissors and they're running around and, you know, they could hurt themselves a little bit along the way. What we're releasing out here in the next week or so is a personal access token capability where you could be scoped directly to what you have access to in the UI. So if you can't do something in the UI, like, you know, manage a particular integration, then that personal access token won't let you do it. So that personal access token will be you can use it for any of our current rest APIs. Manage monitor users can create them, and then you could use it for CLI and MCP. And then we're later going to add OAuth in there as well, which is just a different way to gather a personal access token so that you're scoped to what you have, access to. And then later down in the call, we're gonna get through some of the future enhancements that we've got coming with role based access, but you can kinda see where this is going where you will have, you know, more finer grained permit permissions and privileges that we can give users later on where you can then segment it even further for your personal access token. Yeah. Very cool. I I know we have a demo here, an a demo I actually will walk through with, CLI for, something I was just building myself. So let me share my screen here. Right. So I'm I'm in cloud code here, and I'm I'm just showing the building of a a lookup cache. Right? This is not necessarily something super, technical or taking a lot of time, but what I really liked about this was I'm trying to build a lookup cache for 50 the 50 US states and their abbreviations. So while you're working in the UI would be very simple to to build this lookup cache, this is kind of doing two things. Right? It's building the lookup cache, taking the actual UI actions, but it also has the data around the the states. Right? So I don't have to create a list. So you can see here that I I quickly have the 50 states available with their abbreviations to be used in in a flow. So you could think of, an action that maybe not doesn't take a lot of time in the UI, but maybe takes a lot of research. The CLI is very quickly able to to do that because it has both the context, the knowledge from the element itself, as well as access to perform the commands. And then the next step I'll take is is I'll actually build an entire flow. So this is, certainly more, configuration time than just building a look up cache. But it it'll very quickly go through and choose the commands, and it'll stop and ask me if I want to allow this. Right? So I can have I can always allow or allow once where it can go through. You can see here it's consuming tokens on on, cloud code, but, readily available at no cost for Saleo users. We really see this as a as kind of powering up the user to be able to do more and more on the platform, and we're really excited about. So, I've seen an awesome use cases where it's building agents within flows to accomplish things. It's using the secure connections that I've already established on on on my instance here, so I'm not having to remake, that from scratch. And it's just going through the process, choosing from from the commands, and and building it, for me already. So we'll see here in just a second as it as it finishes up that it'll have the complete flow. It actually added it to a tile that I already had, around Shopify and and Business Central, which is really cool. I didn't even give it those instructions, but it kind of took the context to know to add it here. So I I wanted to add a a flow that synced over my customers. I already had an order flow, which is which is really cool to see with with really just a single line of text that I provided, to to call it using the CLI. Awesome. So we're really excited about all of the the AI, experience that you can have on the platform. But that being said, we're not leaving behind the UI. Right? There's still tons of builders that work in the UI. They do have aura to to upscale that, but we wanna make that builders experience better, you know, every time we do a release to make sure that you can continue to build it in a way that you're familiar with, but do it even faster with more comfort, and build the things that you would like. So Tyler's gonna talk a little bit about, some of the really exciting features that you might be seeing in your flows already that's really just, making day to day kind of operations within the platform, even better. Yep. Thanks, Nate. Yeah. So like Nate said, we're not leaving the rest of the platform behind. And so with that, we've got three really nice enhancements that we've done over the past quarter. One being a lot of flow builder enhancements for how you manage and run flows. So we've we've, fully redesigned flow headers so that it's really consistent, and you have additional functionality that you didn't have before that was functionality that was only available on the back end. So stuff like whenever I manually run a flow, I can choose the exports that you know, if I have multiple exports on the flow, I can choose which exports run at that time. If I run a flow and I have other flows to trigger to run after this particular flow, I can choose whether to run that or not run that. When I run a flow, I could choose a start and end time for, if it's on a delta date. So a lot of enhancements that we historically have already had in the APIs, but we've just completely redesigned that header for flows and given that additional functionality for runtime. And the next would be mapper two dot o. So one of the most critical pieces of functionality for Celigo whenever you're going to an application, right, is you need to map that data to that particular application. And so with Mapper two dot o, we had a lot of user feedback, and incorporated a significant amount of that into its redesign over the past quarter. So with that are things like, larger text fields. When you have handlebar expressions, it's a text area so that you can see the entire handlebar expression without having to, you know, kinda scroll across, without having to click into drawers all the time. You've got little models instead, so you know exactly where you're at in context. Just lets you go through mapper a lot faster than you used to be able to do. And then lastly would be pre parse scripting. This is, somewhat specific for, like, EDI use cases, but can also be for CSVs and JSON files and stuff as well. There's been cases before, especially for EDI files or flat files, where, your trading partner might might be, you know, dumping a bunch of EDI files into one directory, and you maybe don't even know what type of file that that is. It might have the same name or it might just have a time stamp. And until you read that file, you don't know that the file is a eight fifty or a 997. Right? And so you need to do a little bit of preparse scripting to determine, okay. What is this file now that I do know what file type is because I can't parse it out from the name or the directory alone, then handle it, with its appropriate parser or move it to a different directory where I then would handle it with the appropriate parser. So these are just some really three three core enhancements to the product, as we continue working, through the existing UI and existing builder experience. Yeah. Very cool. And without nerding out too much, I'm excited about a new script type. We've had the same same scripts, which are extremely powerful for a long time, but I'm excited for a brand new one, especially around the use case you talked about, Tyler. So, a a small thing here, but not a small thing at the same time. We're always releasing new connectors, especially in the the first, quarter. We have 27 new HTTP connectors. So these are those prebuilt ones you see in the library as well as 45 trading partner connectors to further augment the the b to b EDI offering that we have. Of course, you can always go to the connective and request, connectors that are specific for your organization. We're we're happy to add to our connector library. But, again, just making the build experience even easier, being able to select from our growing library of connectors so that when you're building flows, building APIs, really building anything on the platform, it's it's really that low code, click configure, that that would everybody has enjoyed on the platform. And then I'll hand it over to Tyler, but also a really exciting thing, is is this introducing Celigo tools, which is a brand new kind of object, a first class object on the platform. So, Tyler, if you wanna kind of give a brief overview. of what tools are. Yeah. So if you guys haven't checked out or tried tools yet, it's you know, I I almost see it as maybe the single most important new resource that's legal had in years. Like, it, it's such a new core functionality that's reusable across the entire platform that you historically have not had. And, the way it works is a tool is essentially I say, here's my input that I need for a tool. I can do a ton of encompassing logic, within the tool. So all the stuff you're familiar with, like branching, scripts, transformations, imports, exports along the way are all all within there. And then I have an output node to say, okay. After I've done all this common business logic, here is the standardized output for the thing calling it. And so why it's so important and so reusable is, you know, anything within the platform can call this. You can have AI agents call this. You can have flows call this. You can have tools call tools. You can have, APIs call tools. So any of the standard resources in Celigo you've been using builder canvases for historically can now call tools and tools within tools, to have reusable business logic across all of your flows and across your, entire business. So this really helps with, like, you know, maybe Nate Briant gonna go into this a little bit, but, like, again, just that that reusable logic, that I can use, every everywhere within the platform, and it dynamically changes its runtime based off of whether a flow calls it or whether an API calls it. Yeah. This is a a simple use case, but if you if you've spent time in the platform and have been building lots of flows, you might get familiar with there's there's oftentimes, like, a a step or a piece of logic that you're always rebuilding. So, a lot of ecommerce use cases, multiple channels, you're maybe syncing over customer data. So it doesn't really matter where the customer data is coming from, whether it's Shopify, Amazon, BigCommerce, Walmart. Right? At the end of the day, it's still a customer a customer record. So what I can do now is I can build a tool that is about adding and updating customer data. I know where that customer data needs to land, whether it's my ERP system like Business Central or my database like Snowflake. And I can build that that tool once that is gonna accomplish that business logic, and then I can start invoking it into my flow. So this is a tool invoked in a flow where I have all my different, use cases running and I'm like, I just I can easily add this tool and say, I want I need to add customers. And so you can imagine the scale of this where, I can build basically individual components as tools, and I can just invoke them very quickly. And the great thing about that is when I make a configuration change on the tool, it it obviously is is the configuration for all the instances of the tool. So when a tool gets invoked, if I make a configuration change, it it all of those flows are automatically updated. So, you know, we see businesses make changes to their data structure all the time, or or a business process, and you don't have to go and hunt and find each instance of that. You can just update the tool once. And it really allows you to be consistent with your data, build a more reusable component, and kind of center on, like, a consistent schema or consistent business processes, which is which is really cool. So shifting gears a little bit, off of off of the build components or or the objects here, Tyler's gonna talk about the the kind of evolution of platform user management here, which is which is a really exciting piece, a big change for, how users are gonna be interacting with all the things we just talked about, tools and flows and integration workspaces. So, Tyler, if you wanna talk a little bit about what we've already released and then kind of the layout of the future where we're going with with user management. Yeah. Yeah. So over the past, quarter or so, you know, integrations, workspaces, kinda interchangeably calling them right now, are ways to, you know, segment off who can do what within the platform. Right? So an integration itself is a set of resources that you can then choose who can edit what and who can use and view what within the platform. And so over the past quarter, we've now implemented APIs to be available within, within integration. So, historically, APIs have only been available at the account level, so only administrators could be making APIs. But now anybody that's a manager, custom role user, or a monitor user can see and manage those APIs directly within the integrations that they have access to. And then next would be lookup caches. So lookup caches also have the same thing where they were only available at the account level, or you could create them within mapper, within transformation, but you could never really fully see the lookup cache, table itself. So with this, we've now allowed lookup caches to be registered to integrations just like connections can be registered to integrations. So this allows anybody that can manage that integration. They can now manage that lookup cache or see, see that lookup cache and see the values within there so they can more easily build out the integrations, build out the flows and APIs that reference those. And then coming soon, later in this half would be a lot of, role based access improvements. So, having couple different types of roles allowing you to create custom roles where you can say who can, you know, you know, view what resources in Celigo. Again, personal access tokens, like I mentioned earlier, custom roles and permission sets, and then, user provisioning via SCIM, which, would be really utilized by enterprises to know that when a user leaves the company, they also lose access to everything else. And with SCIM, comes a lot of other things like, you know, now that MCP servers are so prevalent within organizations, MCP servers have long longer living tokens in a different session than the UI. So once that user gets deprovisioned, that's crucial to also deprovision them from MCP servers that they have access to. So all this is coming, later in this half, and exciting stuff to be building out on the governance layer for Sligo. Yeah. And, Tyler, I think what's what's cool about this is if you're an admin already of the platform, you you may not have noticed this stuff. But if if you're an admin looking to add more builders to help you manage Celigo or builders within your org that wanna build automations and use Celigo, it's kinda giving you you the admin confidence, whether it's an IT team or just an admin of this legal platform, that they're gonna have the right access, and the security to say you can build, and I'm gonna give you the right resources. So, really exciting as we look to kind of expand the scope of who who can be a builder and and making sure they have the right access. And and a big part of that, right, is is also a new type of user. So it's not just a a builder, but a very exciting kind of end user. So, Tyler, if you could kinda lay out, like, how people are using it today and then maybe what we're they're gonna see in the future with the with the end user profile. Yeah. Yeah. For sure. So right now, you know, with with all of our users in Celigo, for the most part, they are typically there because they need to, build integrations or they need to monitor integrations or monitor flows. Right? They're fixing errors. They're retrying errors, resolving errors. They're, building out new, you you know, flows and APIs for your business for all your back end office. But when you start releasing MCP servers, which we just put out in January, you run into this other, you know, type of problem where, you know, these users that are accessing MCP servers, they, you know, for the most part, probably are never going to be actually building MCP servers, or they're not going to be building integrations or monitoring errors. So you kind of you end up with two completely different segments of users. You've got your entire company, which would be end users to the MCP server solution that you're building, and then you've got your traditional Celigo users who are the ones building those MCP servers and building those integrations for the back end. And so coming soon, we're releasing a new user type that will be that end consumer, so that consumer of, solutions that are built on the Celigo platform. So for now, it's MCP servers. We have a lot of other things in the road map that would additionally use end users because these are solutions being built on Celigo. And what these end users will be able to do is you'll be able to completely manage them separate from your workspace users, different groups, different users, completely custom roles, just in time provisioning, all that good stuff that enterprises need to manage hundreds of users, and what resources they have access to. Awesome. So we're gonna shift gears here again here. We're gonna talk about EDI, and our b two b manager. So this is, you know, specific for people doing electronic data interchange. Again, you know, though, you'll see a lot of the consistent themes. We wanna make, the platform even easier to use, to manage these type of automations. So the first one we're really excited about is shipping labels. So, for organizations that are doing their own fulfillment on particularly the supplier side where they have their own warehouse, They want to they have to meet requirements from the people they're supplying to to ship pallets, to ship bundles, of of their goods. They can now use shipping labels within their EDI flows to take the advanced shipping notice, apply a a shipping label, and then and send it to a printer. So, we're really excited to be partnering with Bartender here to offer the the label printing solution. So you're gonna have a best in breed, printing solution where you can design and manage your labels, and it'll be, connected easily in the your existing b two b flows or anyone new ones that you build to be able to ship out those, shipping labels to to your printers and and have them generated. There is a a specific app released with with NetSuite around, but this will work with any ERP if you have any fulfillment needs that you're looking to to do with with shipping labels. Another exciting update is around the larger inbound and outbound EDI file support. So now you will have, 50 megabyte limit or four inbound files, and one gigabyte for outbound files. And and we really did this to help support, the larger AI files, of course, but, this is for high volume transactions, with lots of data in in the inbound files. You can you don't have to build any sort of workarounds now within your flows, and you can just continue to see the your your files run through the b two b manager, without having to split it or reprocess any of that data. And and this is especially important. Oftentimes, these requirements are coming from, the trading partners that you're working with, not not yourself. And so we wanna make sure that would you can support any use case, that you would have, to meet the requirements from your from your training partner. So, another update, again, to make the experience better is we just wanted more visibility within, the EDI dashboard and within the EDI flows. So, Tyler, if you wanna just quickly hit on the the acknowledgment visibility, the the new logging features that we have, around EEI. Yep. Yeah. So we've maybe joked a little bit internally that this is the year of observability for Celigo, for all the stuff we've been releasing to give more access to users. And with that would be EDI and b two b manager observability. So, you know, again, historically, acknowledgments, when you're sending acknowledgments to your trading partners to say, you know, I've gotten this file back or I've gotten and processed this file. You didn't have any way to see the acknowledgment that we sent back. So if you ended up in a situation where your the trading partner said they didn't get the file or you needed to see if the file was correctly sent or whatnot, you didn't have really a way to do that without reaching out to our support team. So now we've given you the option to just you can download those and see the acknowledgments that we've sent on your behalf. That way you can troubleshoot yourself, without having to reach out to support. And then next would be a s two MDM logging. Again, kind of the same situation where, you're interacting with a trading partner and you've said you've sent something, and they don't have it, or you can't necessarily see exactly what's being sent. So now you can also see those act see those logs, in the e in the EDI dashboard, so that you don't have to reach out to support and you can fully manage your EDI, flows and integrations all on your own. Yeah. And what with really, like about, these updates, is that that we it EDI is self-service on the on the platform. So we talked to lots of customers that are are frustrated when they have a managed EDI provider where they lack visibility, they lack control. And this is just fully building on that that offering that we have with EDI where you you can really self manage your EDI, which is really critical. Right? It's close to revenue. It's close to how you operate with your trading partners, and so you wanna make sure that you you have that visibility and you don't even now you don't need to reach out to them to see if something went wrong. It's it's it's really, again, more self-service, more control over your business. And so we're maybe saving the best for last, really exciting to to introduce data ingestion here. So I'm really excited about this, as it is a new product. And so we wanted to, do another poll here and kind of understand, how people are using DI, what what what your experience is with it. And, let's we can see the poll results here. Yeah. Awesome. I I think this this is pretty obvious. Right? If if you're going to have any analytics or AI use cases, it's gonna need the data to run. And if it's gonna be making decisions on on behalf, it it needs to have access to it damely, not in a in a timely manner. Right? It can't just be at the end of the day. That's, you know, problems that we in Soligo have have been able to solve as well, right, where, like, you're trying to make reports or or have AI make decisions. It needs to have that data available to it at the time of of run. And so what's what I'll what I'm excited about with this product is that, this is a integration pattern that Celigo customers are are actually already doing on on our platform. So but this release is totally gonna streamline the configuration and management of of that. So in the previously on the platform, if you've been syncing over, in in your data from your sources into your data warehouse, You're building a flow, maybe multiple flows, maybe single flows where you have lots of exports. You're configuring the object type that you're trying to do, and then you're syncing that into a staging table or directly into into your tables. And we have lots of customers that that use for that, be given the flexibility, that they have within the platform. But with data ingestion, we're gonna make the management of that as well as this the building of that even easier. So we will now have a new first class object called a data sync, where instead of having those multiple steps, you're selecting and checking from a list of records. And I don't wanna steal Tyler's thunder because we're gonna show a a a preview here of data ingestion on the platform, but it means less initial configuration. Also really cool is the, built in schema drift. So, when changes are made in those downstream app those upstream applications, excuse me, when something has changed, a field has changed, you it will automatically update within within your data warehouse. So you're not you're spending less time having to manage those flows. It's handled automatically. And what's really exciting is, you know, as the poll kind of showed is is AI business intelligence, it's important that it has that data readily available. But, again, referring back to that big circle, what's what's great about Celigo is it's one platform not just for the ingestion, but also for, like, the reverse ETL or taking the insights that you get out of your data and pushing it back up to the applications, so that the users the business users working in those in those applications can have access to those same insights without having to to leave those applications. And so with Celigo, you can do all of that from a kind of a single plane of control as well. And and that really just means you have faster access to your data. It's less maintenance on both ends, and you can start activating that data faster, whether it's just traditional business intelligence or, those new AI use cases that are coming up. And so now I'll, hand it over to Tyler. I'll I'll share my screen of a a DI walk through so we can see a lot. Alright. Yeah. So what Nate's showing you here is a new data sync that's available, for DI. So what you do is when you go to create a sync, basically, what you you choose a source. So I say I want data from Salesforce in this case. And when you choose Salesforce, you see all of the list of objects that are available within your Salesforce environment. And then you can choose, I want to sync over this particular object, in that case, opportunities, and search for orders, order feed, order history. And so, you know, historically, whenever you built a flow, if you could pause real quick, Nate Briant. Historically, whenever you built a flow, right, you had to create, you know, multiple exports to grab the opportunities. You had to create the SoCal query, right, to fetch, all the fields that were in within that, Salesforce object because, you know, Soquel doesn't let you do, like, select star. Right? And then you had to build that export. You had to build a flow. And so what the sync is doing is just saying, choose your source and then choose the objects that you wanna sync over. So with that, you can go ahead. And then after you choose your source and the objects that you wanna sync over, you choose your destination. So in this case, we're gonna choose Snowflake. Choose a database within your Snowflake environment that you wanna sync this, sync these tables over to. And then at that point, name the schema that we'll be creating underneath that database, in this case, order analytics or Salesforce or Business Central or whatever you wanna name it. And then what we will do is we'll create all of the tables within that within that schema in Snowflake corresponding to all the objects that you've selected to sync over from Salesforce. And then you schedule the sync to run and name it, and then it runs, on a whatever frequency that you would like. And you can you can, on initial run, have an historical sync back to a certain point in time. So, you know, if there's anybody out there that's built these flows, right, to migrate Salesforce data over into a data warehouse, you can pretty easily see how much faster it is to build a sync compared to, you know, a traditional flow, especially when all you wanna do is migrate that raw data over from a source application and into your warehouse to report against. And there's a ton of other functionality in there. We're running a little bit out of time, and we'll probably definitely gonna have another webinar specifically just for DI. But there's a lot of benefits to DI around schema drift policies where if new fields are added, they automatically get added. We maintain the tables for you, if fields are dropped. And we give you all of that visibility and all of that functionality for you to choose what you want and how you want it to sync over. So the really awesome new, offering that we've got here, would be happy. This is kind of how I initially came on the product a couple years ago. It's been, like, my, my forte into product itself. So really excited to see everybody start using it, and see how it does. Awesome. So we'd love for you to share your feedback about this webinar in particular. We've covered a lot of different exciting things. We we changed the format to be a little bit, looking in the past, looking in the future, and we wanna make sure that this is valuable, for everybody, making you, you know, aware of all the great things we're doing, but also getting excited for all the great features coming out that's just gonna further assist you in, being able to to create automations with the within the platform. Awesome. Just wanna call out an exciting webinar that we have coming up. This is, it's particularly around Celigo and Amazon around supply chain services, and and their offering. But what's exciting about this and relevant is that, a big part of this is using AI onboarding or the CLI, the MCP, and Aura to quickly build up entire business processes. So if you wanna see CLI in action around a real use case, especially could be relevant if if you're a a brand worry worrying about how you can meet supply chain requirements, especially in a world where everybody's expecting two day, one day delivery, this is a really exciting webinar that, we're doing with Amazon. So, you'll have speakers and and, specialists there talking about about the offering and then how we, can assist you in in setting that up with with Celigo. Also, feel free to check out the docs. We have some great docs, within the webinar. So, builders hub for more resources. Check out the connective to both put in your request for new connectors as well as, see what other people are building and how they're solving their problems. We have a new AI playbook. We talked a lot about AI today, but, this will help you kind of how you operationalize that, beyond just the discussion of AI. And, also, you know, please register for this series if you found it valuable and and gave your feedback. We'd love to have you back to to talk more about, what we released and what we're releasing in the future. Alright. Awesome, Nate and Tyler. That was that was great. I I know there are a bunch of questions here, so I'm gonna step in to keep us moving through a few questions before we wrap, and I'm going to cherry pick a few. There was a question, can the CLI be used with Claude Code in Versus Code? And the answer is yes, it should be able to. We've used it internally with codex, with Claude Code, with Cursor, so as long as you're able to register the CLI, it should be no problem. Again, check out developer.silica.com for all the documentation on the CLI and the API, and that's where you'll go in the future for details on our platform MCP. So there's another question. Do you need a separate subscription to use Claude Code and so on with the Celigo CLI? The answer is no. Obviously, when you use Claude Code or you use Codex or whatever, you're paying for your own token use, so that part's all on you, and you are free then to use the Celigo command line interface, no additional charge as part of your subscription. There was, I'm jumping around a little bit here, a question about tools, is there ILM support for tools? The answer is no, not quite, not yet. We are, of course, adding that. If you're interested, you may know that or you may be interested to know that if you want to move tools between environments, the CLI plus your favorite LLM is really happy to help you with that. You can easily have the CLI pointing to a prod and a dev instance and just ask to promote or or clone, and it does a really neat job of that. We didn't demo that here today, but I've done that quite a bit myself, including actually moving entire agents and all the accompanying, dependencies. So there was a question that I think has already been answered about the difference between the platform MCP and the Celigo, and MCP servers. Again, just to reiterate the point, MCP servers are what you create yourself to expose whatever services you have, via MCP and make them available to either Soligo AI agents or to external LLMs, whereas the platform MCP here, we are talking about using your favorite LLM to control the platform itself. So just like the CLI allows you to, you know, control the platform, you know, maybe create users or modify the permissions, create integrations and flows and APIs and so on, PlatformMCP is exactly like that. Essentially the similar scope but focused on control of the platform through the LLM using natural language as opposed to requiring a developer experience. There were a question about Aura, one related to Aura and legacy sandboxes. Right now, yeah, in in older pricing models, Aura doesn't we we disabled it for legacy sandboxes. We are working on migrations, so stay tuned on that. There are other customer communications, but Aura, of course, can work quite happily in production and will provide feedback to the user, so it shouldn't be making changes without your permission. The final one that I'll take and maybe ask Tyler, related oh, Yeah. go ahead. was just gonna add, if you want to be migrated earlier, into multi environment, you can request from your account rep, and we can see if that's possible to do, sooner. Yeah. And and then, Tyler, while you're there, there was a question about Aura and NetSuite, hooks, whether it's possible for Aura to help there. And I didn't know the answer, but I figured you might. Yeah. If you have execution logs enabled, then or would be able to see the execution logs, that happened within NetSuite against that script. The only probably downside is that we wouldn't actually be able to, see that script itself, like, the contents of how that script is going. So you might have to, like, copy that script into Aura itself, and then it could help you troubleshoot. But it should at least a combination of those two would be able to help you out. There was also a question related to retailer specific labels, and and Bartender does have a library of, of templates and so on, but also provides a designer that makes it easy to, to extend, right, to add custom labels because I know sometimes there are, peculiarities that need to be, contended with. So I think that's about all we have time for and and the bulk of the questions. Anything we've missed here, I'm just keeping an eye on the clock. I want a excellent time. I wanna thank our our presenters for today, Nate and Tyler, for putting on a great show. Thank you all for attending. We've had a great deal of questions, and we look forward to continuing the conversation. You can always meet us at the community website, connective.solego.com, or wherever you, come to get your Celigo information, and and thanks all for your your time and attention today. Appreciate it. Thank you.