Progress Software Corporation ($PRGS)

Hello, and welcome to today's webinar, a journey to high-performance multi-database connectivity. Before we begin, I have a couple of housekeeping slides to show you the webinar control panel and how you can participate. [Operator Instructions] And one final note, we are recording this session, which will be available in the coming days. So let's start by quickly introducing to you our speakers for today. Today, we are joined by our speakers, Dennis Bennett and Imran Hussain. Dennis has over 20 years of experience in working with companies to help solve their data connectivity challenges. In the past, he's worked as an application developer, DBA and QA engineer. At Progress Software, he is the Principal Sales Engineer for the data direct data connectivity and integration products. Imran is an experienced hands-on software architect involved in the architecture, design and implementation of WinSQL and other software published by Cinematrix. He has been in the software industry for the past 25 years and has significant experience and expertise in application integration and enterprise architecture. Imran has a bachelor's degree in computer and electrical engineering and a minor in economics from Rutgers University. I'll now turn it over to Imran to start the presentation.

Thank you very much, Rachel. Like Rachel said, my name is Imran, and I am a Senior Manager here at Cinematrix. We have been using DataDirect and WinSQL together for the past, I would say, probably 10 to 15 years. And it's -- the partnership has been great ever since we started, both for our customers because when our customers use WinSQL, the drivers to connect to databases come with it. And therefore, this combination has been very, very good. So, let me get started with the slides and towards the end, I'll do a demo showing how WinSQL can connect to different databases. So WinSQL is a querying tool. You can submit queries and review the results and all that kind of stuff. I am going to focus more on stuff that you may not find in other applications that are in the same realm. So I'm going to go through the -- some of the features, and then I will talk about some of the special features. So like any querying tool, you can run queries, you can view catalog and you can fetch the metadata from the back end. It includes many wizards that allow you to submit SQL queries for insert update, all that kind of stuff, the administrative tasks. You can do importing and exporting. So you can import data from one database and export to another database and things like that. And it is tightly integrated with the Windows scheduler. So many customers use WinSQL to schedule their tasks. So you can write queries or importing or exporting data and then you can run it through a scheduler. Now some of the unique features in WinSQL most querying tools either work with ODBC or JDBC. But in WinSQL, both technologies are supported natively. When you connect, you pick if you want to use an ODBC driver or a JDBC. Other unique feature is heterogenous queries, meaning you can actually do -- let's say, you have a query going to DB2, but then some of the data comes from an Oracle database. You can do something like that in WinSQL. You can do backups and restored across databases. So for example, you can back up a Microsoft SQL server database using WinSQL and then restore it into a DB2 database or an Oracle database, things like that. Another unique feature in WinSQL is dynamic executables. WinSQL can generate stand-alone executables that can run without having WinSQL installed. All you need is the connectivity to the database and the ODBC driver. And then any person who has access to these executables can execute them. And this is very useful for managers who don't have technical knowledge how to run queries, but all they do is just double-click an Exe file and the results of the queries are there. Additionally, the latest version of WinSQL is integrated with AI. So right from the WinSQL's interface, you can submit AI queries, you can chat with AI, and it can give you answers based on the design of your database. It has the ability to draw ER diagrams. So with a just a mouse click, you can get a visual overview of your database. And then lastly, it has a feature to share comments, meaning one team member can enter a comment to a table and the other team members can view those comments on those tables and fields and indices that are in the database, which is also very, very useful. So now let's talk about integration with DataDirect. Like I said, for the past maybe 15 years, WinSQL has been shipping ODBC drivers from DataDirect. And in the latest version of WinSQL, we will be shipping JDBC drivers as well. So since these drivers are part of WinSQL, there's no additional install that you need. All you need to do is simply point to your database and you can get going. There are no client-side libraries. So this is one of the major advantages of using DataDirect. I'll give you an example of DB2. If you want to use an ODBC driver from IBM, it usually requires that you install the DB2 client libraries on the machine and then configure the libraries, then the ODBC driver talks to that library, which then connects to the database. When using DataDirect, that middle layer is not there. And therefore, not only the connections are a little bit faster, but you have less software to install. And of course, the DataDirect drivers support the latest specs. If there's any problem that you run into, it's very easy to reach the support for both DataDirect, as well as Cinematrix. So now let me get started with the demo. So here, in this example, I am going to connect to 2 databases. The first will be IBM DB2 and the second one will be Microsoft Access. And I'll show you why I am connecting to DB2 and Microsoft Access. So first, I'm connecting to DB2, and you can see I'm using JDBC for that. So I'm going to connect. Let me move this over and maximize this. Second, I am going to connect to a Microsoft Access database. So here are the list of all the databases that I have. The ones that are configured with JDBC have JDBC in front of it and the ones that are with ODBC have ODBC in front of it. Now in case of Microsoft Access, it is using an ODBC driver. So now I'm connected to Access and DB2. Like any other querying tool, you can write queries, you can view the catalog. So for example, if I type select all from employees and I misspelled. Okay, it's employees. All right. So here's the result of the query. The primary keys are in red, the foreign keys are in blue. And if I click any of the primary keys, I can bring data from a related table here. Now let me show you a few things that are very specific to WinSQl. So for example, I have my catalog here. In the catalog, I have this employee table. I click the manage relationship and it draws the entire relationship for me. If I want, I can add, remove tables, move the tables around and things like that. One other thing that I can do is since I have a connection to Microsoft Access open in this other tab, I can simply drag and drop my object from DB2 to Microsoft Access. And when I do that, it actually opens up the export wizard. And here, I can type I want my table in to be employee. And if everything is okay, so this is the design of my new table in Access, which did not exist before this import. So I'm going to do next. These are all my columns. If I need to change anything, I would do here. But right now, everything is good. So I'm just going to click next, next, and then finally, I'm going to finish it. And it says that it export 42 rows into Microsoft Access. So now I go to Microsoft Access. And right now, I probably won't have that table there. So the employee table is missing because I just added it. So now I need to refresh the catalog. And once the catalog is refreshed, I see my employee table. And if I browse the data, I see this data that actually came from IBM DB2. There are many features here under the tools menu that do all sorts of things. For example, it can actually create code connecting to your database tables, so you can create codes in either Java or C-Sharp. You would do this data access object. You can do scheduler, you can do online, off-line backup where you can back up one database, but then restore that database into a different RDBMS. So you back up DBS, but the data actually gets restored into Oracle. Now when you are doing this, you can only do the restoration of the data. If you have like stored procedures and trigger then, you have to handle those manually. So that is pretty much it as far as the demo goes. Now of course, due to the time limitations, I cannot go through every feature that is available here. But if anybody is interested, we have several videos on our website. And of course, if you have any questions, I'll be more than happy to answer them towards the end of the session. So from -- so that's said, I'm going to ask Dennis join, and he's going to be talking about how DataDirect works with WinSQL.

Thank you, Imran. That was a very important to see. And I know looking at your website, and you mentioned it briefly that you guys are doing more around AI to kind of add value to WinSQL. Can you speak to in any more details as far as what you guys are doing with AI to kind of bring value to your customers?

Absolutely. And thank you very much, Dennis, for bringing that up. So yes, let me spend a few minutes, actually probably less than a few minutes to talk about AI. So in WinSQL, in the latest version, there is this tab here called AI chat. And you can configure different providers, including Google, ChatGPT or a lot of those providers. And then you can ask questions related to your database. Now right now, let me connect your DB2 because it has a little bit more tables. So right now, like I'm connected to DB2 and when I do in my AI chat, I can ask questions to AI that are related to my tables. And I have an option of using whichever LLM I decide to use. I can do this configuration and pick the LLM, so you can see there are different LLMs here. Right now, I've selected Ollama, meaning the AI provider is actually running locally on my network. So that -- it takes care of all the privacy issues associated with AI. So -- and now here, I can ask any questions related to my table. So for example, if I don't know how to join 2 tables, I can ask AI to do it. Or I can simply ask AI, hey, please go through my table design and tell me what this database is doing or how do I write certain queries. I can even write queries and then ask -- in this case, it's a very simple query. But if this was a very complicated query, I could write click and I can select hey, explain with AI. And then AI will try to figure out what the query is doing and give you a very nice explanation of what the query does. So Dennis, does that answer your question?

It does. I appreciate it. All right. Thanks, Imran. That was great. Just to speak a little bit about Progress DataDirect, what we do. As Imran said, we are kind of helping provide connectivity to many different back-end data sources and do it in a way just to make it easier for software vendors like Cinematrix to be able to connect to a lot of different sources without having to be on the hook to provide the connectivity themselves or put it on their customer to go ahead and download drivers and get going. So we've been doing this for a very long time, and we have a lot of customers similar to Cinematrix and WinSQL. We go from the very largest software vendors all the way down to start-ups who are bringing new products to market and want to kind of have data connectivity in the box. Okay. So kind of moving on, as I said, we've been doing it for a long time. We go all the way back to the foundations of ODBC and JDBC and OData. And we stick very closely to those specifications. And one benefit of doing that is it makes it easier for software vendors like Cinematrix to be able to support multiple different data sources without having a different code path for DB2 versus Oracle or something else. These sources are going to behave in much the same fashion. We will support the ODBC specification, the JDBC specification. You saw when Imran displayed WinSQL where you were able to see all the metadata, see the schema, see the tables, drop down, drill down, see the columns and the data types and all that. And those were all metadata calls made under the covers by WinSQL and DataDirect was just responding as far as what was there on DB2 and so forth. And it would be the same if you were even using like a driver to a Software-as-a-Service application, like, say, for example, like a salesforce or something like that. So we've been doing this for a long time. And most of what we do is working with other software vendors like Cinematrix to add value to their products and to help them delight their customers. So as far as like, as I mentioned, why did people do it? Well, people like Cinematrix with WinSQL, they do what they do best. Imran was showing just a lot of the different things you can do. And you could see there was a lot more things their product could do to make it easier for users to examine data sources and to get value from them. So, we're able to help them do that in a way that provides value and also to make it where less technical users are able to get value. Imran mentioned doing things to set up for maybe a less technical manager. We find that by including the drivers in the product, there is less friction and proof of concepts. You don't put the onus on the less technical users to find the right drivers and drop them in and configure them. It just makes the POCs go smoother. It makes it easier to onboard new customers. Sometimes we kind of talk about it as batteries included. This makes life very much easier for the user and also makes it easier for Cinematrix as they are supporting their customers. So -- the other thing we do is, we provide very responsive support to our partners in this. Sometimes we kind of laugh about it and say it's one throat to choke. But if it is one vendor you're going through for multiple different sources, you know who to talk to, you're used to working with us, and we do everything in our power to make you successful. So in effect, DataDirect would be the driver factory for these different products like WinSQL. Kind of going on, we're able to work, as I mentioned, with software partners from the smallest startups to the very largest companies. And so we're able to be flexible and make it make sense, basically make it a win-win where we're both happy and successful. And hopefully, we grow together, start small and as our partners get successful, they have more usage, and it works very well for all of us. We also work with direct end users. We're not focusing on that as much today, but all the benefits I talked about are there for the end users also for corporate users. And this is just a list of the data sources we support. We started on the left side in the relational space, expanded into Software-as-a-Service and cloud sources as well as big data and cloud data warehouses and such. But the benefit is, we make them all behave the same. So if WinSQL is doing metadata calls against IBM DB2, they could do the same metadata calls against an SAP S/4HANA or something like that and it would behave the same way. And the other thing is WinSQL embeds a lot of our drivers, but there is also -- WinSQL gives the capabilities to use external other drivers and beyond that. And I've done that myself with WinSQL. I've actually been a user for a number of years. As Imran demonstrated, I realized there's a lot of functionality I wasn't taking advantage of, which I'll have to kind of go back and do it. But I've really found value in the product. And I often will drop in some of our other drivers just to kind of test it that way. I find it very useful for that. I'll just quickly mention another product of ours, the Autonomous REST Connector just because I showed you that last list of sources, and it's a big list, but the truth is there are many, many other sources and many of them expose the rest API. And with the Autonomous REST Connector, we're able to put a kind of a SQL template in front of it. We're not creating a second copy of the data, but we're basically translating SQL calls. And I've used Autonomous REST Connector with WinSQL before. So if I put in a SQL statement, we're able to translate that into a rest request, take within Autonomous REST Connector and then turn into relational results for that. So definitely something we're seeing more and more usage of these days. And we've built a number of models already. So you can use these as templates to kind of get started with the Autonomous REST Connector, but you could also -- if you've got your own REST source that maybe it's internal or something like that, we're able to allow you to model it using the Autonomous REST Connector Composer. And we're also doing some work around AI also to kind of make it even easier. That's always the benefit is to just make things less friction and help people get more value out of their data. And I'll speak to one last product is hybrid data pipeline. That is -- most of what we do are driver-based stuff, ODBC and JDBC drivers, but we also have a server-based product, hybrid data pipeline. Some of the benefits of that are you could connect to a single source and single source connect to hybrid data pipeline and be able to get to about 40 or so different data direct drivers that are built within the product. And we see this often with Software-as-a-Service vendors who have maybe a data warehouse behind it that they want to give their customers access to, whether it's from a tool like WinSQL or whether it's from a BI tool to do data visualizations. But hybrid data pipeline kind of makes it easier to handle that and do it in a way where there are -- you can set up limits and throttling, you can do auditing to see who's connecting and what they're requesting as far as data. So just another product just kind of thrown out there. And finally, we just have a lot of resources out there. You're certainly welcome to go look. We have a number of blogs and tutorials as well as a progress YouTube channel with a lot of DataDirect kind of resources out there. And I know Imran mentioned they have a bunch of videos also, which I've definitely seen out there. So again, I think that's all I kind of want to talk about. I really appreciate Imran joining us and kind of showing how our products work together. And I think at this point, I'm going to throw it back to Rachel and see if there are any questions out there.

Thanks, guy. All right. So let's go ahead and turn on your cameras for the live Q&A. And just a reminder to our audience, please ask your question for Dennis and Imran at the questions panel. So let's get started. What do you recommend ODBC or JDBC and are there any differences in terms of how WinSQL can act?

So the answer really depends on what are you using, what technology are you using for your own back-end systems. From a functional standpoint, there is no difference. Underneath, of course, there are 2 complete technologies, but the -- when you are submitting queries and getting results from the database, there's no real difference. So it really depends on what you are using on the back end. Now often, companies use both technologies. And when they use both technologies, they want to make sure that the queries they write are first tested. So WinSQL is a great tool for that -- for those kind of tasks, right? You want to test all your queries, make sure everything works and then, of course, use those queries in your actual applications. So in that case, you will be using both technologies. But as far as one versus the other, they are pretty much the same. There is no difference in terms of performance or anything like that. You write the same queries in JDBC as well as ODBC. I hope that answers your question?

Yes. Okay. Dennis, what do you mean when you say that data track can be the driver factory for an independent software vendor?

Yes. When I say Driver factory, I mean that Progress DataDirect is actually providing all the connectivity and kind of taking that task away from the software vendor where they don't have to build and maintain drivers for every database or every data source their customers would potentially want to use. So, it just makes it easier, especially with a new product where you can have a lot of connectivity from the start rather than having to kind of slowly build up and support that. So -- and these things change, too. New databases, new data sources come about, new platforms, cloud platforms, new APIs or certainly evolution of existing ones, and we're able to kind of handle those updates and certify the drivers with them. So it kind of takes away a lot of the responsibility for connectivity from the engineers at a company like Cinematrix and kind of puts it on us, which is that is our function and has been for a number of years. So they can kind of rely on us and helps them to add value to their product. Most software products that connect to data, the more different data they can connect to, the more value they provide. So we're able to kind of provide that value.

Great. And for a software company supporting many customer environments, how does DataDirect help deliver broad and consistent connectivity? And what does that do to the vendor support load?

Yes, I can take this. So I guess a big challenge would be for a software vendor is every customer's environment is different, different databases, different versions, different platforms. Some data is cloud, some is on-prem. And what we provide is a kind of a consistent layer. I mentioned that we adhere closely to the ODBC and JDBC specifications. So the drivers are going to behave in much the same way regardless of the back end. So, it just makes it where -- I mentioned like a single code path and also not something like worrying about it works here, it doesn't work there kind of thing. So it just makes it easier to support different drivers, different behaviors with the kind of standardized connectivity through a single partner like DataDirect, yes.

This one for Imran. Does WinSQL allow migration of data from SQL server database tables to host for us SQL database table?

Yes, absolutely. So it doesn't matter what is the source and what is the target. Basically, what it does is you first connect to your source database and then you write your SQL queries that will -- either SQL queries or you say, hey, I need to export so and so tables. In either case, it creates what we call internally a data bag. A data bag is basically a holder for the data. And then those data bags can be restored either in the source database or any target database. Usually, WinSQL takes care of all the data types. So if in one source, the data type is CHAR, but in the other one, it's VARCHAR, it takes care of those things. Sometimes you do run into situations where the target does not have the data type. For example, XML is one of those examples where some of the database support an XML type, but the other ones don't. In that case, you have to make certain decisions. Maybe you want to translate those XML into a VARCHAR. So there might be some minor adjustments. But yes, basically, you can back up any database or any table and then restore it to any other destination.

I am currently using DB2 with IBM's driver and often run into issues when setting up a new machine. I find the WinSQL driver easier to set up. Can I use the WinSQL driver with our in-house application? If yes, how?

Okay. So the drivers that come with WinSQL, they are published by DataDirect. However, the licensing that is available for those drivers is only for WinSQL. They are what we call a closed driver. If you need to use those drivers because you feel that, hey, those are a lot easier to use, then you need to contact DataDirect, their sales teams and download a driver from DataDirect and then use those drivers. But yes, a quick answer is, you can use the drivers that come with WinSQL with WinSQL, but not with any other application.

Yes. What Imran is describing is -- and this is something we do when we work with software partners is we give them the ability to, first of all, white label it. So, any messages could potentially come out, not saying DataDirect, but saying a Cinematrix, WinSQL, but also to lock it for use with their application. So especially in the case of like JDBC, ODBC too, where they couldn't just grab it and use it with any other tool. It kind of protects, I guess, both us and the partner that the product is being used in the way intended.

Yes. And WinSQL is a great tool to try drivers from DataDirect because those drivers are already built in and all you got to do is just give it a shot and see if they work. If they work, then yes, you go at the DataDirect directly and get those drivers for your application.

And when using the AI features, does WinSQL connect to Cinematrix server?

No. So when you use AI, you pick a provider, if, whichever provider you pick, that's the company that you're sharing your private information with. For example, Google Gemini is very, very popular. If you decide to use Google Gemini and if you -- and of course, when you use Google Gemini, there are 3 levels of what you want to share. Level 1 is you don't want to share anything. Level 2 is -- or actually Level 3 is that you share everything. And then Level 2 is in between the 1 and 3, right? You only share data that is needed. So for example, if you have a customer called -- if you have a table called customer. But in your query, you type find me clients living in New York. The AI is smart enough to know, hey, the person is asking for client, but I have a table named customer. It's not called client, right? So in that case, the schema for that customer table is going to shared with Google Gemini. Now you also have an option of using a local LLM, like I mentioned in my demo. So Ollama is a great tool for that. If you use Ollama inside your network, you can use the Ollama API with WinSQL. And in that case, none of your private data goes to either Google or whatever LLM you use, everything stays private. So that privacy is very, very important, and you decide how you want to deal with it.

Dennis, can you expand more on the differences between the DataDirect DB2 driver and the native driver?

Yes. And actually, Imran kind of mentioned some of it. So the DB2 drivers, the native drivers are quite good. They're usually more environment specific. You'll have LUW, Linux, UNIX Windows, you'll have iSeries and you'll have zSeries, and they're all kind of separate. So, one thing we do differently is in the DataDirect driver, we use the wire protocol. So a single driver can connect to DB2 on any platform, also without having to go through a server, like in some cases, there might be like a DB2 Connect server if you're using a native driver. And the other thing is what Imran mentioned is often some of those drivers would require client libraries, which is a little bit more difficult to deal with. Since we're using the wire protocol, we don't need any client libraries in the driver. And we're able to, as I said, connect to all the different major DB2 platforms in kind of a simple and consistent fashion.

Imran, can you elaborate more about how backup and restore works across different databases? Can I back up a DB2 database and restore it to MS SQL server?

Yes. So this is similar to the question about the gentleman asked earlier, right? Yes. So answer is yes, you can definitely back up. I'm sorry, did you say DB2 and SQL Server, right? Yes. So there are 2 ways to do it. Either you write a select query in your DB2 and then you export the results of that query to Microsoft SQL Server. That's method #1. Method number two is, you selectively pick the tables that you want to export and then you say, I want those tables in the other database or you can back up the entire database and then restore it on the other end. Now when you back up entire database and if you have like hundreds of tables, it might become a little bit tedious because, let's say, if the data types don't match, you will have to match them manually, right? But if it's just from DB2 to, let's say, SQL Server, I would say probably 95% of the data types are matching between those 2. So you won't have to do anything. But if you are taking 2 sources that are completely different, that's where some of the frictions come in. And in that case, you just have to -- when it asks you, hey, is this create table statement correct? You just modify the create table statement. So instead of, let's say, the XML data type on the target database, it becomes stream or our bar card.

And what information is that to be LLM related to my database when using AI?

Yes. So this is again a similar question, right, the privacy issue. So, you have 3 levels. You can either decide to not share anything about your design. If you decide to do that, you'll still share, hey, what type of database you're using, what is the version of the database and things like that, but that -- none of that is private. And the third level is that you share every design of every table in your database, okay? And that's a good option if you're trying to get like a summarized view or maybe a design question that you have related to your database, right? You could ask, hey, my current design does XYZ, but I need to change this design so that I can do these additional tasks. That's where sharing your entire database comes in handy because then the LLM can look at all the tables, make sense of all the tables and then give you an answer. The second level is somewhere in between #1 and #3, which is hybrid. And what it does is before it sends the queries to the back end, it is going to first ask, hey, this is a question, give me a list of tables that are relevant in this case. So like I said, a client, it's very common that people are calling it client, but on the database, it's actually customer. But the LLM is smart enough to know, hey, there's a relationship between clients and customers, right? So it can say, okay, you need to share the design of the customer table. So all those things, the privacy settings are all on for you and you decide how much information you would like to share with the LLM. And of course, if you use Ollama, which is the local, then you don't share anything.

Great. This one is for Dennis. What changes when a software company relies on a single vendor for connectivity versus managing multiple drivers?

I guess the main thing is it would simplify things with a single vendor instead of multiple. Really, you just have, I said, like one throat to choke. You have one place to go and you don't have people pointing fingers at another vendor or whatever. So we are kind of responsible for your connectivity. And typically, that means faster resolution of any issues and just a good support model. The support we're able to provide is one of the reasons people like to partner with us. It just makes life easier for the software vendor. We have 24/7 support. You can open cases via phone, through a portal. We can -- basically, hopefully, by providing better support for the software vendor, you'll have fewer escalations, hopefully, fewer issues, happier customers. It just simplifies matters, just kind of having one place to go for data integration.

And speaking of support, what are DataDirect support policies?

Yes. So it's 24/7. I don't have the service level objectives in front of me, but that is something -- it's global support. We have offices in different time zones. We can do follow the sun on very escalated cases and such. So yes.

Great. It looks like we have time for one more question. For privacy, if I use Ollama, will Ollama still call public LLM to answer my query? And if it does, how to make sure it does not share my information?

So no, that's the benefit of Ollama, right? When you run Ollama, you don't need Internet. In fact, I often run Ollama on my laptop. I don't need the Internet. I can ask all the queries that I have. And I can actually submit queries using APIs. So none of that need an Internet connection. So yes, Ollama is great in that respect.

Great. Thanks, guys. Well, if you have any more questions for Dennis or Imran, you can get in touch with us at Progress. Thank you so much for joining our webinar, and I hope you all have a great day.

Thank you.

Thank you very much. Bye.